Báo cáo khoa học: Factors involved in the assembly of a functional molybdopyranopterin center in recombinant Comamonas acidovorans xanthine dehydrogenase pot

Factors involved in the assembly of a functional molybdopyranopterin
center in recombinant
Comamonas acidovorans
xanthine
dehydrogenase
Nikolai V. Ivanov
1
, Frantisek Huba
´
lek
2
, Manuela Trani
1
and Dale E. Edmondson
1,2
Departments of
1
Chemistry and
2
Biochemistry, Emory University, Atlanta, GA, USA
Previous work from this laboratory has shown that the
spectral and functional properties of a prokaryotic xanthine
dehydrogenase from Comamonas acidovorans show some
similarities to those of the well-characterized eukaryotic
enzymes isolated from bovine milk and from chicken liver
[Xiang, Q. & Edmondson, D.E. (1996) Biochemistry 35,
5441–5450]. Therefore, this system was chosen to study the
factors involved in the expression of functional recombinant
enzyme in Escherichia coli to provide insights into the
assembly of the functional Mo-pyranopterin center. Genes
xdhA and xdhB (encoding the two known subunits of the

native enzyme) and putative genes xprA and ssuABC were
sequenced. Heterologous expression of the xdhAB genes
in E. coli JM109(DE3) produced active enzyme. The Mo
content was 0.11–0.16 mol per ab protomer, while the Fe
and FAD levels were at stoichiometries similar to that of the
native enzyme. The XDH activity increased sixfold when the
culture was grown under conditions of low aeration
(6 LÆmin
)1
) as compared with high aeration (12 LÆmin
)1
).
Co-expression of the xdhAB genes with the Pseudo-
monas aeruginosa PA1522 (xdhC) gene increased the level of
Mo incorporated into the expressed enzyme to a 1 : 1 stoi-
chiometry. Under these conditions, high levels of functional
protein (2.284 UÆmg
)1
and 8.039 mgÆL
)1
of culture) were
obtained independently of the level of culture aeration.
Therefore, the assembly of a functional Mo-pyranopterin
center in XDH requires the presence of a functional xdhC
gene product. The purified, recombinant XDH shows
spectral and kinetic properties identical to those of the native
enzyme.
Keywords: xanthine dehydrogenase; Comamonas acidovo-
rans; prokaryote; molybdopyranopterin; FAD.
Xanthine oxidoreductases have been extensively investi-

gated because of their physiological and medical importance
[1], serving as a prototype for the study of multiredox
centers in an enzyme system. They belong to a group of
molybdenum hydroxylases that catalyze the hydroxylation
of substrates using solvent water as the oxygen source.
Hydroxylation of hypoxanthine to xanthine and of xanthine
to uric acid is considered to be their major biological
function in purine catabolism, with either NAD
+
[for the
xanthine dehydrogenase (XDH) form] or O
2
[for the
xanthine oxidase (XO) form] acting as electron acceptors.
XDHs are found in both prokaryotes and eukaryotes,
with the enzymes isolated from bovine milk, chicken liver
and Rhodobacter capsulatus being functionally and struc-
turally the best characterized [2,3]. These enzymes contain
one FAD, two [2Fe)2S] centers, and one molybdopyra-
nopterin monophosphate (MPT; the site for substrate
hydroxylation), per subunit. A cyanolyzable terminal sulfur
ligand on the Mo center [4] is absolutely required for
catalytic activity.
With the recent determinations of the crystal structures of
both a eukaryotic (from Bos taurus [5]) and a prokaryotic
(from R. capsulatus [6]) XDH, detailed mechanistic probes
of the function of the Mo center in the hydroxylation
reaction would benefit from site-directed mutagenesis
studies of recombinant XO/XDH. The detailed molecular
mechanism of substrate hydroxylation and other questions

have remained unanswered owing to a lack of progress in
the expression of functional eukaryotic XDHs at high levels.
Heterologous expression of recombinant Rattus norvegicus
XDH in a baculovirus system [7] resulted in a predominantly
Correspondence to D. E. Edmondson, Department of Biochemistry,
Emory University School of Medicine, Rollins Research Center,
1510 Clifton Road., Atlanta, GA 30322, USA.
Fax: + 1 404 727 3452, Tel.: + 1 404 727 5972,
E-mail:
Abbreviations: IPTG, isopropyl thio-b-
D
-galactoside; MPT, molyb-
dopyranopterin monophosphate; XDH, xanthine dehydrogenase;
XDH
AB
, recombinant XDH expressed in Escherichia coli NI453
(without xdhC); XDH
ABC
, recombinant XDH expressed in
Escherichia coli NI850 (with xdhC); XO, xanthine oxidase.
Enzymes: xanthine dehydrogenase (E.C. 1.1.1.204) and xanthine
oxidase (E.C. 1.1.3.22); Comamonas acidovorans XdhA (Q8RLC1)
and XdhB (Q8RLC0); Bos taurus XDH (P80457); Rhodobacter
capsulatus XdhA (O54050) and XdhB (O54051); Rhodobacter capsul-
atus molybdopterin cofactor insertase XdhC (Q9 · 7K2); Pseudo-
monas aeruginosa XdhA (Q9I3I9), XdhB (Q9I3J0) and XdhC
(E83456); Escherichia coli putative molybdenum cofactor sulfurase
ycbX (P75863); Escherichia coli putative molybdenum cofactor inser-
tase yqeB (Q46808); Escherichia coli Elongation Factor EF-Tu
(P02990).

Note: nucleotide sequence data are available in the GenBank database
under the accession number AY082333 Version: 11 August, 2003.
(Received 11 August 2003, accepted 9 October 2003)
Eur. J. Biochem. 270, 4744–4754 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03875.x
nonfunctional demolybdo form. Attempts to express the
recombinant rat enzyme in the baculovirus system [8,9] were
reported to produce  10% functional enzyme at low
expression levels. Expression of Drosphila melanogaster
XDH in Emericella nidulans [10] produced similar results,
yielding unstable recombinant enzyme with low activity.
Homologous expression of D. melanogaster XDH mutants
in D. melanogaster produced minimal levels of the enzyme
(< 10% of native) [11,12]. Given the complexity of the
MPT center in XDH, it is not surprising that expression of a
fully functional enzyme has remained a formidable chal-
lenge. Successes have been achieved in the expression of the
molybdopyranopterin enzymes sulfite oxidase and dimethyl-
sulfoxide reductase in Escherichia coli [13,14]. These Mo
containing enzymes do not have the complication of the
terminal sulfur ligand on the Mo center as a requirement
for activity. However, the successful expression of these
enzymes in E. coli suggests the potential for success in
expressing a recombinant XDH in this organism.
Expression of XDH requires the incorporation of the
MPT center, placement of the terminal sulfur ligand on the
Mo, and the incorporation of two [2Fe)2S] centers and an
FAD into the recombinant enzyme. Therefore, the host
organism has to be able to provide these cofactors and the
means for their incorporation into the expressed protein in
order to form an active enzyme.

Proteins functioning in the sulfuration of the molybdo-
pyranopterin cofactor have been identified in D. melano-
gaster [15,16], Homo sapiens [17], B. taurus [18],
Arabidopsis thaliana [19], and E. nidulans [16], and are
homologous to prokaryotic NifS-like proteins, which are
involved in Fe–S cluster biogenesis [20]. A similar protein,
ycbX,alsoexistsinE. coli [16], suggesting that this
bacterium has the machinery to sulfurate the MPT in a
recombinant XDH. Genes similar to the recently identified
R. capsulatus xdhC (which encodes a 33 kDa protein with
proposed function of either a MPT chaperon or a MPT
insertase) [21] have also been identified in a number of other
prokaryotes (Fig. 1), including a weakly homologous gene,
yqeB,intheE. coli genome. However, no apparent
homologues in eukaryotes have been observed.
Previous work from this laboratory has shown that the
XDH isolated from Comamonas acidovorans exhibits the
same cofactor content and many spectral and kinetic
properties similar to those of the eukaryotic enzymes [22].
Rather than an a
2
homodimer of  300 kDa, as observed in
eukaryotes, the bacterial enzyme is an a
2
b
2
heterotetramer
of subunits exhibiting molecular mass values of  60 kDa
(b-subunit) and  90 kDa (a-subunit). As a good deal of
information is available in the literature on the C. acidovo-

rans XDH, attempts of its expression appeared to be a
worthwhile effort which, if successful, would provide a
system permitting the use of site-directed mutagenesis to
probe the functional roles of amino acid residues implicated
in the catalytic mechanism of this class of enzymes.
In this article we describe the sequencing, cloning and
conditions required for the high-level expression of func-
tional recombinant C. acidovorans XDH in E. coli.Itis
shown that coexpression of the Pseudimonas aeruginosa
xdhC gene is required for stoichiometric MPT incorpor-
ation into the recombinant protein and for expression of
fully functional enzyme. While this manuscript was under
review, Leimkuhler et al. [23] reported the successful
expression of R. capsulatus XDH in E. coli. The similarities
and differences of the results of their studies with those
reported here are discussed.
Materials and methods
Reagents, strains and vectors
All reagents and medium components were purchased either
from Sigma-Aldrich Co. or from Fisher Scientific Co. E. coli
JM109(DE3) was a epst from M. W. Adams (University of
Georgia, Athens, GA, USA). E. coli strains TOP10 (Invi-
trogen Inc., Carlsbad, CA, USA), XL-1-Blue or XL10-Gold
(Stratagene Inc., La Jolla, CA, USA) were used for routine
transformations. Vectors pGEM-5Zf(+) (Promega Co.,
Madison, WI, USA) and pCR2.1 (Invitrogen) were used for
routine cloning, and vector pET23a(+) (Novagen Inc.,
Madison, WI, USA) was used as the expression vector.
C. acidovorans (ATCC 15667) was maintained on minimal
medium plates, as described previously [22].

DNA purification and manipulation
C. acidovorans plasmid preparations were performed using a
combination of cell lysis with SDS and equilibrium centri-
fugation in CsCl/ethidium bromide gradients [24]. Purifica-
tion of cloning and expression vectors for use in DNA
sequencing was performedusing the QIAprep Spin Miniprep
kit (Qiagen Inc., Valencia, CA, USA). All DNA manipula-
tions were carried out using standard protocols [24].
LASERGENE
software (DNAstar Inc., Madison, WI, USA)
was used for sequence manipulation and assembly,
BLAST
(NCBI, :80/BLAST/) was used
to identify gene homologies, and the Windows version
of
CLUSTAL X
(NCBI) and
BOXSHADE
(Swiss Institute
of Bioinformatics, />BOX_form.html) were used to generate multiple protein
sequence alignments. Promoter regions were predicted
using the NNPP server (itfly.org/seq_tools/
promoter.html).
Fig. 1. Gene organization of several sequenced prokaryotic XDH. Most
prokaryotic xdh operons contain the xdhC gene, encoding a molyb-
dopterin cofactor insertase, located downstream of the xdhAB genes.
The known exceptions are Berkholdaria mallei and Berkhol-
daria pseudomallei. In these organisms, the xdhC gene is either absent
or located elsewhere in the genome. C. acidovorans, Comamonas
acidovorans; P. aeruginosa, Pseudomonas aeruginosa; R. capsulatus,

Rhodobacter capsulatus; E. coli, Escherichia coli.
Ó FEBS 2003 Influence of xdhC on functional XDH expression (Eur. J. Biochem. 270) 4745
Peptide sequences of XDH
Large-scale fermentations (200 L) of C. acidovorans were
performed at the University of Georgia Fermentation
Facility. Enzyme purification from frozen cell paste was
performed using a simplified method, described previously
by Xiang & Edmondson [22].
N-terminal sequence analysis was obtained by automated
Edman degradation of purified XDH blotted onto
poly(vinylidene difluoride) membranes, or of peptides iso-
lated by HPLC at the Emory University Microchemical
Facility using an Applied Biosystems 494cLC Protein
Sequencer. A ReflexIII MALDI-TOF (Bruker Daltonics,
Billarica, MA, USA) and API3000 triplequadruple (Applied
Biosystems, Foster City, CA, USA) mass spectrometers
were also used for peptide sequence determination.
Gel electrophoresis
SDS/PAGE and native PAGE gels [25] were used to
confirm the identity of recombinant proteins. Western blot
analysis was performed using rabbit polyclonal antisera
raised against C. acidovorans XDH, as described previously
[22]. Proteins were detected by staining with Coomassie
Brilliant Blue (Sigma), or by activity staining in the presence
of xanthine/Nitro Blue tetrazolium [26]. Low molecular
weight SDS/PAGE standards (Bio-Rad Laboratory Inc.,
Hercules, CA, USA) and a 1 kb ladder (Promega) were
used, respectively, as protein and DNA standards.
Cloning and sequencing of the
C. acidovorans xdh

genes by PCR
Eleven polypeptides, sequences of which were obtained by
Edman degradation of C. acidovorans XDH peptides and
are shown in the multiple alignments (Fig. 2), were chosen
to produce degenerate primers. DNA sequencing and DNA
synthesis were performed at the Emory University Micro-
chemical Facility.
The degenerate primers were used in a stepdown PCR
method, described previously [27], together with 2.6%
dimethylsulfoxide and 1
M
betaine as additives. The PCR
products were extracted using the MinElute Gel Extraction
kit (Qiagen), ligated and then transformed into E. coli
TOP10 competent cells according to the TOPO TA cloning
kit (Invitrogen) manual. The required clones were identified
by digestion with EcoRI, sequencing and assembly into a
continuous sequence. To extend the sequence from both
ends, random primers were generated based on the repeti-
tion frequency of all octamers in the sequenced portion of
the C. acidovorans xdh operon using the program
OPTFREQ
,
written in this laboratory in ANSI C and available from the
authors upon request. Four of the highest frequency
octamers (gcaaggcc, cgagctgg, gtggcgca, gcctgcat), with a
GC content close to that of the C. acidovorans genome
(68%), were used to generate 3¢ termini of the PCR primers.
The 5¢ termini of the PCR primers were synthesized at equal
nucleotide concentrations. The second primer was derived

either from the known 5¢ end of the xdhA gene
(ggcaggaattgaatgcag) or the known 3¢ end of the xdhB gene
(gcccagtacctacaagattc). Localization of xdhAB genes on the
C. acidovorans plasmid was established using the AlkPhos
Direct System with the 360 bp probe generated by PCR
from the forward (5¢-tccatcattcatgacgacc) and reverse
(5¢-atgtacggctccgtcttcct) primers and C. acidovorans plasmid
DNA as a template.
Activity and protein assays
XDH activity was measured spectrophotometrically by
monitoring the production of uric acid at 295 nm
(e
295
¼ 9600
M
)1
Æcm
)1
) in activity buffer (100 m
M
Tris/
HCl, 1 m
M
EDTA, 0.6 m
M
xanthine and 2 m
M
NAD
+
,

pH 7.8) at 25 °C. One unit of XDH activity is defined as the
amount of enzyme catalyzing the production of 1 lmol of
uric acid per minute at 25 °C. The Biuret [28] or the Bearden
[29] assays were used to estimate protein concentration. The
protein concentrations of purified enzyme samples were
determined by measuring the absorption at 450 nm
(e
450
¼ 37 000
M
)1
Æcm
)1
). The level of functional enzyme
refers to XDH species containing a functional Mo catalytic
center. The functionality is estimated as a ratio of change in
the absorbance at 450 nm after anaerobic reduction of the
enzyme with 1 m
M
xanthine relative to the absorption
change after reduction by dithionite.
Expression of recombinant
C. acidovorans
XDH
AB
in
E. coli
The xdhAB genes, containing coding sequences for both
a-andb-subunits of C. acidovorans XDH, were amplified
Fig. 2. Location of the xdhAB gene operon on an isolated Coma-

monas acidovorans plasmid. Agarose gel (A) and Southern blot (B)
analyses of 0.3 lgofpurifiedintactC. acidovorans plasmid (lane 2)
and of 0.3 lgofplasmiddigestedwithSacII (lane 3). Lane 1 contains
0.5 lg of a 1 kb DNA ladder standard. The Southern blot was
developed using a 360 bp probe, as described in the Materials and
methods. Besides being a DNA size marker, lane 1 serves as a negative
control, showing that the probe is specific for xdhAB genes. Lane 2
demonstrates that the probe highlights the band corresponding to the
C. acidovorans plasmid. Lane 3 of the gel (A) shows that after incu-
bation of the plasmid with SacII, it becomes completely digested;
however, on the blot (B) it remains as a sharp band of lower molecular
size. Lane 4 shows the effect of double digestion of 0.3 lgofC. acido-
vorans plasmid with NdeI/SacII, which corresponds to a fragment of
the plasmid containing the xdhAB gene.
4746 N. V. Ivanov et al. (Eur. J. Biochem. 270) Ó FEBS 2003
using stepdown PCR [27] mediated by the forward primer
xb101+ (5¢-gccgcccatatgcaccaccaccaccaccacagcaccagtca
gaactct), the reverse primer xa100– (5¢-gtggtgaattcagc
cagtgtgcccttg), and pNIall2 as a template. To facilitate
purification of recombinant enzyme, the xb101+ primer
incorporates an encoded hexahistidine tag into the 5¢ end of
the xdhA gene. The reaction product of 4.2 kb was purified
and ligated into a pET23a(+) vector, producing the
pNI453 expression vector. Correctness of the construct
was demonstrated by DNA sequencing. The expression
construct, pNI453, was introduced into E. coli
JM109(DE3) using electroporation (400 lLcuvettes,
2 mm gap, 2500 V,  5 ms), according to the manufac-
turer’s manual. The resulting E. coli strain, NI453, was
cultured overnight (16 h), at 21 °C in LBANG media

(Luria–Bertani medium supplemented with 100 mgÆL
)1
ampicillin, 1 m
M
guanosine and 0.25 m
M
sodium molyb-
date), without isopropyl thio-b-
D
-galactoside (IPTG) induc-
tion. For large-scale preparations, two 12-L fermenters
(New Brunswick Scientific, Edison, NJ, USA) were used
for overnight culture (16 h) of E. coli NI453 at a stirring
rate of 400 r.p.m. with an airflow of either 6 LÆmin
)1
for
low aerated preparations or 12 LÆmin
)1
for highly aerated
preparations. The cells collected by centrifugation were
stored at )80 °C.
Preparation and expression of XDH
ABC
, containing the
gene encoding
P. aeruginosa xdhC
,in
E. coli
NI850
The xdhC gene (PA1522) was amplified from genomic

DNA of P. aeruginosa PAO1-LAC, via stepdown PCR,
using forward xc1+ (ctgaacaagcttgatcgggaggatgacgag) and
reverse xc1e– (gcggggctcgagtcaggattcgtgggcgc) primers. The
PCR product was digested with HindIII/XhoI, purified,
ligated with T4 ligase into plasmid pNI453 and transformed
into E. coli JM109(DE3) by electroporation. Clones con-
taining the correct inserts were identified by restriction
digests with HindIII/XhoI and by DNA sequencing. The
growth of E. coli NI850 was performed under conditions
similar to those described for E. coli NI453, except that high
aeration was achieved at 6 LÆmin
)1
and low aeration at
2LÆmin
)1
airflow in a 10 L BIOFLO 3000 bioreactor (New
Brunswick Scientific).
Purification of recombinant XDH
AB
and XDH
ABC
Cell pastes of E. coli strains NI453 and NI850, cultured as
described above, were resuspended in three volumes of
buffer A (50 m
M
KH
2
PO
4
,1m

M
EDTA, 2 m
M
2-merca-
ptoethanol, pH 7.8) and disrupted using a combination of
lysozyme treatment (0.1 mgÆmL
)1
lysozyme incubated at
37 °C for 20 min) and sonication (five 1 min intervals at an
80% power level of 550 Watts).
The cell extract was clarified by centrifugation at 45 000 g
for 30 min, mixed with 0.8 volumes of DE-52 resin and
packed onto a column of 0.4 volumes of fresh DE-52 in a
batch manner. The column was washed with buffer A until
the absorbance at 280 nm (A
280
) was lower than 0.1; the
washing buffer was then replaced with buffer B (50 m
M
KH
2
PO
4
,2m
M
2-mercaptoethanol, pH 7.8). After washing
with five column volumes, the protein was eluted with a salt
gradient (0–1
M
NaCl) in buffer B. Fractions containing

catalytic activity were pooled and applied directly to a Ni
2+
bound Talon column.
To remove nonspecifically bound proteins, the Talon
column was washed with buffer C (50 m
M
KH
2
PO
4
,
300 m
M
NaCl, 5 m
M
imidazole, 2 m
M
2-mercaptoethanol,
pH 7.8) until the A
280
was lower than 0.03. Bound material
was eluted with buffer D (50 m
M
KH
2
PO
4
,300m
M
NaCl,

150 m
M
imidazole, 2 m
M
2-mercaptoethanol, pH 7.8). The
active fractions were pooled, concentrated by ultrafiltration
in an Amicon concentration unit (Millipore, Billerica, MA,
USA), and chromatographed on a Sephacryl S-300 (Amer-
sham Biosciences) gel filtration column equilibrated with
buffer E (50 m
M
Hepes, 1 m
M
EDTA, 2 m
M
2-mercapto-
ethanol, pH 7.8), to exchange the buffers from D to E,
concentrated by ultrafiltration and stored at )80 °C. The
recombinant enzymes purified from E. coli strains NI453
and NI850 were named XDH
AB
and XDH
ABC
, respect-
ively,
Protein characterization
UV/Vis spectra were measured using a Lambda 2 spectro-
meter (Perkin-Elmer) at 25 °C using buffer E as a reference.
CD spectra were recorded in 10 mm path length cells on an
AVIV 62DS spectrometer. EPR spectral data were recorded

using a Bruker ER-200D spectrometer equipped with
an Oxford Instruments cryogenic system for liquid He
temperatures.
Metal and cofactor analysis
Prior to the metal and cofactor analysis, protein solutions
were passed through a small Chelex-100 column to remove
all adventitious metals. Mo analysis was performed using
the dithiol method [30], while iron analysis was performed
by using the modified ferrozine complex method [31]. Metal
analysis of the recombinant protein samples were also
performed using Thermo Jarrel-Ash965 inductively coupled
argon plasma spectroscopy at the Chemical Analysis
Laboratory, University of Georgia. The flavin content of
recombinant protein samples was measured by UV/Vis
absorption spectra after acid precipitation of the protein
[32].
GenBank accession number
The DNA sequences of xdhAB, xprA and ssuACB are
available in GenBank (accession number AY082333).
Results
Gene localization and sequence
The observation of a naturally occurring plasmid as part of
the C. acidovorans genomic DNA [26] raised the question of
location of the xdhAB operon. The plasmid size is estimated
to be 24 kb, by comparison to the mobility of a supercoiled
DNA standard on electrophoresis (data not shown).
Agarose gel (Fig. 2A) and Southern blot (Fig. 2B) analyses
of the intact and of SacII digested C. acidovorans plasmid
DNA provide direct evidence that the xdhAB genes are
located on the plasmid.

Ó FEBS 2003 Influence of xdhC on functional XDH expression (Eur. J. Biochem. 270) 4747
Amino acid sequences obtained by Edman degradation of
tryptic or Lys-C generated peptides of purified C. acidovo-
rans XDHwereusedtodesignprimersaccordingtothe
alignment of the peptide sequences with B. taurus, D. mel-
anogaster,andR. norvegicus XDHs (Figs 3 and 4). The
stepdown PCR, with addition of dimethylsulfoxide and
betaine, was used to overcome the high GC content
( 68%) typical of the Comamonas genus. The resulting
sequence assembled from sequencing reads of the PCR
clones covered both xdhAB genes separated by one adeno-
sine nucleotide, as determined from the alignment of N- and
C-terminal protein sequences from both subunits; their
translated amino acid sequences are shown in Figs 3 and 4.
All peptide sequences (Figs 3 and 4), identified by Edman
degradation, and by MALDI and ESI mass matching, were
located on the protein sequence that corresponded to the
xdhAB nucleotide sequence. According to the sequence
obtained, the b-subunit contains 535 amino acids with a
calculated average molecular mass of 57 752 Da and the
a-subunit contains 808 amino acids with a calculated average
molecular mass of 87 392 Da. The total calculated average
molecular mass of the heterotetrameric (a
2
b
2
) protein is
290 298 Da. Electrospray MS of purified C. acidovorans
XDH showed the a-subunit mass to be 87 427 ± 30 Da
and the b-subunit mass to be 57 774 ± 15 Da (data not

shown). The differences between the observed and calcula-
ted masses, 35 Da for the a-subunit and 22 Da for the
b-subunit, are within the experimental error of the instru-
ment. These data also demonstrate that the enzyme does not
contain any detectable post-translational modifications.
Four additional genes, with homologs in P. aeruginosa
and E. coli genomes, encoding putative xanthine/uracil
permease (xprA) and three subunits of the putative sulfo-
nate ATP-binding cassette transporter (ssuACB), were
identified by extension of the 3¢ terminal sequence of the
xdhAB genes. None of these genes were similar, at
the protein level, to the xdhC gene, which is a member of
the majority of bacterial xdh operons (Fig. 1). Prediction of
a prokaryotic promoter in front of xdhA andinfront
of ssuA suggests that xprA is cotranscribed with xdhAB
genes as a single operon. The presence of a predicted Shine–
Dalgarno motif for xdhB in front of the xdhA stop codon
implies translational coupling between the two subunits. A
similar relationship is observed between xdhABC genes in
P. aeruginosa and R. capsulatus.
Expression of recombinant
C. acidovorans
XDH
AB
and XDH
ABC
in
E. coli
Optimal expression levels of the functional enzyme XDH
AB

in E. coli requires the presence of 0.25 m
M
Na
2
MoO
4
and
1m
M
guanosine in the media, no IPTG induction, and a
low (6 LÆmin
)1
) air supply. Guanosine is added because it is
a known precursor in the biosynthesis of the MPT cofactor
and is permeable to the cell. Induction of XDH
AB
expression with IPTG concentrations ranging from 0 to
1m
M
results in decreased specific activities in cellular
extracts.
The influence of aeration levels on XDH functionality
was tested because the enzyme catalyzing sulfuration of the
Fig. 3. Sequence alignment for the small subunit (xdhA). The sequence of the small subunit of Comamonas acidovorans was aligned with those of the
XDHs from Bos taurus [38] and Rhodobacter capsulatus [3]. XDH peptides identified by Edman degradation (solid arrows), or by MALDI or ESI
mass matching (dotted arrows), are indicated. The peptides used to generate PCR primers are underlined with double arrows.
4748 N. V. Ivanov et al. (Eur. J. Biochem. 270) Ó FEBS 2003
Mo cofactor has been shown to utilize cysteine as the source
of sulfur [19]. Therefore, the level of sulfuration of
recombinant XDH may be influenced by the redox state

of the thiols in the cell. Indeed, the level of XDH
AB
activity
observed in crude cell extracts was  60-fold higher in cell
cultures grown with a lower level of aeration (6 L of air
min
)1
) than that observed with a higher level of aeration
(12 L of air min
)1
)(Table1).
Under optimal conditions, functional XDH
AB
was
observed at a level (0.25 UÆmg
)1
of total protein and
2.16 mg of enzyme L
)1
of culture) comparable to that
produced by C. acidovorans grown on minimal media
containing hypoxanthine (0.46 UÆmg
)1
and 1.87 mgÆL
)1
)
[22] and lower relative to the reported expression of
R. capsulatus XDH
ABC
in E. coli (12.5 mgÆL

)1
)[23].As
we were unable to isolate xdhC from C. acidovorans,agene
encoding xdhC was cloned from P. aeruginosa. Its coex-
pression within the xdhABC expression construct resulted in
the expression (2.28 UÆmg
)1
and 8.04 mgÆL
)1
) of functional
C. acidovorans XDH
ABC
in E. coli cellular extracts. It also
showed that the xdhC gene product is active in the assembly
of an XDH from a different bacterial species. In contrast to
the results obtained with the expression of XDH
AB
in E. coli
(see above), the level of culture aeration had no influence
on the level of expression of observed XDH
ABC
activity.
Purification and characterization of XDH
AB
and XDH
ABC
XDH
AB
and XDH
ABC

were purified, as described in the
Materials and methods, and visualized on Western blots
(Fig. 5A), confirming the cross-reactivity with antisera
raised against C. acidovorans XDH. Activity staining with
xanthine/Nitro Blue tetrazolium of native gels containing
equal levels of activity (10 lU) of recombinant and native
enzymes showed no detectable difference in the mobility of
intact functional enzyme (Fig. 5B). All expressed enzyme
samples migrated as a single band on native PAGE and
as two bands on the denaturing gels (Figs 5A and 4C),
Fig. 4. Sequence alignment for the large subunit (xdhB). The sequence of the large subunit of Comamonas acidovorans was aligned with those of the
XDHs from Bos taurus [38] and Rhodobacter capsulatus [3]. XDH peptides identified by Edman degradation (solid arrows), or by MALDI or ESI
mass matching (dotted arrows), are indicated. The peptides used to generate PCR primers are underlined with double arrows.
Ó FEBS 2003 Influence of xdhC on functional XDH expression (Eur. J. Biochem. 270) 4749
corresponding to the two subunits of  60 and  90 kDa,
respectively.
Although Western blot analysis showed that both XDH
subunits are expressed at comparable levels, the expression
levels of neither the XDH subunits nor XdhC were
sufficiently high to be detected in crude cellular extracts by
staining with Coomassie Brilliant blue (data not shown).
SDS/PAGE gels (Fig. 5C) of purified XDH
AB
and
XDH
ABC
showed a co-purified third band at  40 kDa
(the band is very insubstantial in the XDH
AB
preparation).

MALDI-TOF MS analysis identified this protein to be the
E. coli elongation factor EF-Tu (43 kDa). The presence of
EF-Tu in the XDH preparation is not expected to affect any
of the kinetic parameters, or spectral or physical properties
of the purified enzyme, because the XDH concentration is
calculated based on the absorbance at 450 nm contributed
by the [2Fe)2S] centers and the FAD cofactor.
The specific activity of the purified XDH
AB
grown under
conditions of high aeration was 67-fold lower than that
grown under conditions of low aeration (Table 1). Both
enzyme preparations contained the appropriate levels of
Fe–S and FAD cofactors, as judged by their respective
visible absorption spectra (Fig. 6) and by metal and flavin
analyses. Both preparations contained low levels of Mo
Table 1. Effect of culture aeration on the functionality and cofactor content of pure recombinant Comamonas acidovorans xanthine dehydrogenase
(XDH) expressed in Escherichia coli NI453 (XDH
AB
) and NI850 (XDH
ABC
).
Property XDH
AB
XDH
AB
XDH
ABC
XDH
ABC

C. acidovorans
preparation [22]
Low aeration High aeration Low aeration High aeration
Specific activity, UÆmg
)1
11.18 0.1673 46 70 50.00
Functionality, % 15 0–1 78 93 90–100
A
280
: A
450
ratio 8.4 6.1 11.2 8.2 5.5
FAD:ab
a
1.0 1.1 1.2 1.2 1.0
Fe:ab
a
3.5 3.9 2.9 2.7 3.9
Mo:ab
a
0.11 0.16 1.1 1.1 1.1
k
cat
,s
)1a
27 ± 1 0.4 ± 1 118 ± 1 179 ± 4 120 ± 3
K
m
xanthine, l
M

66 ± 3 67 ± 2 85 ± 2 86 ± 5 68 ± 5
K
m
NAD
+
, l
M
148 ± 4 156 ± 4 100 ± 5 113 ± 8 163 ± 3
a
Based on the moles of ab protomer determined from the absorption at 450 nm and e
450
¼ 37 000
M
)1
Æcm
)1
.
Fig. 5. Electrophoretic characterization of recombinant XDH. Western
blot analysis (A) with 0.25 lg of protein loaded per lane, (B) native
PAGE activity staining with 10 lU of enzyme, and (C) a Coomassie
Brilliant blue-stained SDS/PAGE gel with 5 lg of protein loaded per
lane, compares purified XDH
AB
(lane 1), XDH
ABC
(lane 2) and native
Comamonas acidovorans XDH (lane 3). The third band of  40 kDa
observed in C is identified as Escherichia coli EF-Tu and is copurified
in both recombinant preparations (more clearly visible in the XDH
ABC

preparation).
Fig. 6. UV/Vis absorption spectra of pure enzymes. (A) Native Com-
amonas acidovorans xanthine dehydrogenase (XDH), (B) recombinant
XDH
AB
purified from Escherichia coli NI453 expressed at low aer-
ation, (C) recombinant XDH
AB
purified from E. coli NI453 expressed
at high aeration and (D) the enzyme XDH
ABC
isolated from E. coli
NI850 expressing the xdhABC construct (similar for either aeration).
The absorption spectra have been offset from one another for clarity of
presentation.
4750 N. V. Ivanov et al. (Eur. J. Biochem. 270) Ó FEBS 2003
(10–15% of a stoichiometric level). Therefore, the similar-
ities of metal and flavin content for both preparations
suggest that the difference in activity is directly linked to the
difference in the level of sulfuration of their respective MPT
centers. We estimate that the enzyme preparation purified
from cells grown under aeration conditions of 6 L of air per
min has all of its Mo centers in an active, sulfurated form,
whereas the enzyme purified from cells grown under
aeration of 12 L of air per min has less than 2% of its
Mo centers in an active, sulfurated form.
It was previously shown that R. capsulatus xdhC is
required for efficient incorporation of MPT into R. capsul-
atus XDH [21]. Similarly, it appears that the coexpression of
xdhC with the xdhAB genes is required for the expression

of functional recombinant C. acidovorans XDH (termed
XDH
ABC
) with a stoichiometric level of MPT. Growth of
this strain under aeration at 6 LÆmin
)1
, and purification of
the expressed XDH
ABC
, resulted in an enzyme preparation
with a stoichiometric quantity of MPT (Table 1) as well as
of FAD and Fe–S centers and a level of functionality
approaching 100%. Growth of the E. coli cultures under
different aeration conditions did not significantly affect the
functionality of the recombinant XDH
ABC
(Table 1).
These results provide direct evidence that the gene
product of xdhC functions in E. coli as either MPT or Mo
insertase, as suggested previously, by the work of Leimkuh-
ler et al.[21],onR. capsulatus XDH expressed in R. cap-
sulatus. The protein product of xdhC appears to exhibit a
redox role in maturation of the XDH
ABC
by eliminating the
dependence of its functional expression on the redox state of
the cell (Table 1). These data also show that P. aeruginosa
xdhC is able to participate in assembly of the C. acidovorans
XDH in E. coli.
The steady state kinetic properties of the purified

recombinant C. acidovorans XDH
ABC
are essentially iden-
tical (Table 1) to those determined previously for the
naturally occurring enzyme [22]. Therefore, the recombinant
enzyme appears to be functionally identical to the native
enzyme.
Spectroscopic studies of purified XDH
AB
and XDH
ABC
The UV/Vis spectrum of recombinant C. acidovorans
XDH
AB
(Fig. 6), purified from a 6 LÆmin
)1
aeration
preparation of E. coli NI453, is very similar to the spectrum
of native enzyme in either an oxidized (as isolated) or a
reduced (with 0.6 m
M
xanthine) state. In a range of
300–500 nm, the Fe–S clusters exhibit spectral properties
typical of a [2Fe)2S] ferredoxin absorption maxima (315,
420, and 467 nm), which overlap with FAD absorption
maxima at 370 and 450 nm [33]. As observed with bovine
and native C. acidovorans XDHs, the absorbance in the
300–500 nm range is bleached upon addition of 1 m
M
xanthine owing to reduction of the Fe–S and FAD centers,

and it is reduced even further upon addition of dithionite
crystals, indicating that the enzyme exhibits  15% func-
tionality (Table 1). Similarities of the oxidized and reduced
absorption spectra of the recombinant XDH, with those of
the native XDH and of bovine XDH, are consistent with the
metal and flavin analysis data (Table 1) and demonstrate
the presence of two nonidentical [2Fe)2S] iron–sulfur
centers and an FAD cofactor.
The identity of the two [2Fe)2S] iron–sulfur clusters was
further confirmed by CD (Fig. 7) and EPR (Fig. 8)
spectroscopies. Previous data in the literature have reported
that the visible CD spectrum of bovine milk XO is
dominated by contributions from the two [2Fe)2S] centers
[33], a property also exhibited by C. acidovorans XDH [22].
Figure 7 shows the visible CD spectra of the recombinant
XDH
AB
in its oxidized form, after reduction with xanthine,
and after reduction with dithionite. The overall shape of the
oxidized spectrum was identical to that of native C. acido-
vorans XDH and similar to that published [34] for bovine
milk XO, showing a characteristic positive band with a
maximum at 430–440 nm and two negative bands with
maxima at 370–380 nm and at 560–570 nm. Similar CD
spectra were obtained for purified preparations of XDH
AB
(grown at low aeration) and with XDH
ABC
(independent
of the culture aeration level). The level of reduction by

xanthine was higher than expected from the level of
functionality (15%), owing to the reduction of non-
functional enzyme by functional XDH in the time required
for collection of six CD scans. These spectral data further
support the fact that recombinant XDH contains [2Fe)2S]
clusters identical to those in the naturally occurring enzyme.
Low temperature EPR spectra of reduced samples of
purified recombinant preparations (Fig. 8), taken at 10 K
and 70 K, demonstrate the presence of [2Fe)2S] I and the
fast-relaxing [2Fe)2S] II (observed at temperatures lower
than 20 K) in both forms of recombinant enzyme (XDH
AB
and XDH
ABC
). Thus, the incorporation of these two Fe–S
clusters within the enzyme is independent of the coexpres-
sion of xdhC. The EPR spectrum of XDH
ABC
at 70K shows
Fig. 7. CD spectra of recombinant xanthine dehydrogenase (XDH)
AB
(low aeration) in oxidized (solid line), xanthine reduced (dashed line) and
dithionite reduced (dotted line) states. TheCDspectrumforXDH
ABC
was identical (only one for the oxidized form is shown, in large dots).
Upon addition of xanthine, the reduction was incomplete, as seen by
the further reduction with dithionite, and indicates the presence of
nonfunctional enzyme. The level of reduction by xanthine was higher
than expected from the level of functionality, as a result of the
reduction of nonfunctional enzyme by functional XDH in the time

required for collection of six CD scans. Conditions: 22.89 l
M
of
XDH
AB
(low aeration) and 5.4 l
M
of XDH
ABC
(low aeration) in
50 m
M
Hepes, 1 m
M
Tris,pH 7.8,25°C. Excess 1 m
M
xanthine and a
few crystals of dithionite were used for reduction.
Ó FEBS 2003 Influence of xdhC on functional XDH expression (Eur. J. Biochem. 270) 4751
asignalat 349 mT caused by Mo(V), which is not
apparent in the spectrum of XDH
AB
at 60K because the Mo
content is  10-fold lower.
Discussion
Our knowledge of the structure and function of
Mo-dependent enzymes has increased dramatically over
the past several years as a result of successes in the structural
elucidation of a number of enzymes in this class by X-ray
crystallography. This structural information can be exploi-

ted to use site-directed mutagenesis, in a detailed manner, as
additional probes of enzyme structure and mechanism. The
complexity of the molybdopyranopterin site (the catalytic
site for most of the enzymes in this class) adds further
difficulty in the successful expression of molybdoenzymes in
the xanthine oxidoreductase family.
The most important factor involved in producing a
functional MPT center in recombinant XDH, using the
E. coli expression system, is the coexpression of xdhC
(thought to be an MPT insertase [21]). Failure to coexpress
this protein leads to a recombinant enzyme preparation
being highly dependent on the level of culture aeration or
redox state of the cell and having a low Mo content
(Table 1), consistent with the results of Leimkuhler et al.
[21] on the expression of R. capsulatus XDH
AB
in R. cap-
sulatus where xdhC was disrupted. The data presented here
show that Fe–S cluster formation and FAD incorporation
into the recombinant XDH do not require a functional
XdhC.
In contrast, Leimkuhler et al. [23] find the level of MPT
in their R. capsulatus XDH preparation, expressed in
E. coli, to be stoichiometric and independent of the presence
of xdhC, and their data suggest that the level of MPT
sulfuration is lower in the absence of xdhC than in its
presence. Assuming that the level of incorporated MPT and
Mo is the same in an enzyme preparation, and that the
R. capsulatus XDH is very similar to C. acidovorans XDH,
we suggest the following explanation for these seemingly

contradictory results.
Partial (10%) incorporation of molybdenum into recom-
binant C. acidovorans XDH expressed in E. coli
JM109(DE3) may be explained by the presence of an
intrinsic E. coli Mo insertase. It is possible that a product of
the E. coli gene yqeB, exhibiting weak homology to the
P. aeruginosa xdhC or R. capsulatus xdhC, can perform this
function with a broad specificity for MPT and MGD
cofactors. However, as the majority of MPT produced by
E. coli JM109(DE3) is converted to MGD by MGD
synthase [35], only a small amount of MPT is available
for incorporation into the expressed apoprotein. On the
other hand, E. coli TP1000, used for the expression of
R. capsulatus XDH [23], carries a disrupted MGD synthase
and produces MPT as a final product of the cofactor
biosynthesis, resulting in its stoichiometric incorporation
into R. capsulatus XDH. From the results presented in this
study, one would predict that the expression of the latter
enzyme, under low culture aeration conditions, might yield
a fully functional enzyme in the absence of xdhC.Alternat-
ively, the intrinsic E. coli Mo insertase may have a lower
insertase activity with the heterologously expressed
C. acidovorans XDHthanwiththeR. capsulatus enzyme.
The co-expression of P. aeruginosa xdhC with C. acido-
vorans xdhAB,orR. capsulatus xdhC with R. capsulatus
xdhAB, results in the production of enzymes with stoichio-
metric levels of MPT, independently of E. coli or of aeration
Fig. 8. EPR spectra of dithionite reduced recombinant Comamonas
acidovorans xanthine dehydrogenase (XDH)
AB

(low aeration) and
XDH
ABC
preparations. All spectra were calculated as an average of
four or five scans to reduce the signal-to-noise ratio. The xanthine
reduced spectrum for XDH
AB
wasveryweakasonlyasmallfraction
of [2Fe)2S] centers were reduced by the substrate. Xanthine-reduced
spectra for XDH
ABC
were similar in shape and intensity to the
dithionite-reduced spectrum (data not shown), as the enzyme is fully
functional. The molybdenum signal at 349 mT was more prominent in
XDH
ABC
(at70K)comparedwiththeXDH
AB
preparations
.
(A) The
spectra were taken at 12 K (top) and 60 K (bottom). The enzyme
concentration was 81.7 l
M
in buffer E; microwave frequency
9.65 GHz; microwave power 0.2 mW; modulation frequency
100 KHz; receiver gain 2 · 10
4
; modulation amplitude 1 mT. (B) The
spectra were taken at 10 K (top) and 70 K (bottom). The enzyme

concentration was 70 l
M
in buffer E; microwave frequency 9.66 GHz;
microwave power 0.2 mW; modulation frequency 100 KHz; receiver
gain 2 · 10
4
; modulation amplitude 1 mT.
4752 N. V. Ivanov et al. (Eur. J. Biochem. 270) Ó FEBS 2003
levels, indicating that XdhC is specific for MPT insertion. It
remains for future work to delineate the detailed function
and mechanism of the xdhC gene product in MPT insertion
into the molybdoenzymes.
The finding that XDH
AB
activity levels are dependent
on the level of culture aeration suggests that, in the
absence of XdhC, the incorporation of the terminal sulfur
ligand into the Mo center is dependent on the redox state
of the cell [23]. Our functionality data suggest that at a
lower culture aeration level, all the molybdenum present
in the XDH
AB
preparation is sulfurated by an E. coli
sulfurase, which may be different from the E. coli MPT
insertase and which has a function that is redox state
dependent. A multitude of information has accumulated
over the past few years to demonstrate that the enzyme
catalyzing this reaction contains pyridoxal phosphate and
provisions for a persulfide bond formation [16]. Appar-
ently, this enzyme is present in E. coli and is functional

with the heterologously expressed XDH. Current evidence
supports the source of the sulfur moiety to be free cysteine
and that it is required to be in the reduced form to be an
effective sulfur donor [19]. We propose the following
explanation for the effect of the level of culture aeration
on the expression of recombinant protein. Raising the
cellular redox potential by increasing aeration levels is
expected to reduce the level of cysteine and increase the
level of cystine in the cell, thus reducing the availability of
a sulfur donor for the sulfuration reaction producing
active XDH. This would explain the dependence of
activity levels of XDH
AB
observed on culture aeration.
How the presence of XdhC protects against this effect
remains to be established, but does suggest a redox role
for this protein in addition to its insertase function.
It is of interest that expression of the MPT-dependent
enzyme rat liver sulfite oxidase in E. coli [36] does not
require the coexpression of an MPT insertase, as there was
no mention of a dependence of the level of culture aeration
in the expression of recombinants. These results, together
with the results presented in the current report, demon-
strate that the incorporation of MPT into XDH is a more
complex process than the incorporation of MPT into
sulfite oxidase. Differences have been observed in the
reactivity of MPT in bovine milk XO compared with that
of rat liver sulfite oxidase [37]. The authors propose that
bovine XO contains MPT in the quinonoid dihydro form
and that the MPT in sulfite oxidase is a different dihydro

isomer. These suggested differences in the respective MPT
cofactors of these two molybdoenzymes may be related to
differences in requirements for MPT insertion into the
apoproteins.
Little is known regarding the structure or mechanism of
the gene product of xdhC. It appears to be present as part of
the xdh operon in a number of prokaryotes (Fig. 1). A
screen of the genomic DNA library of C. acidovorans has
been performed using a probe based on the consensus
sequence of the xdhC genes present in other bacteria and, to
date, has not been successful in identifying xdhC. Our failure
to locate the xdhC gene within the xdh gene operon of
C. acidovorans suggests that it resides at another location, as
it appears to be essential for molybdopterin incorporation
into XDH. Unfortunately, XdhC proteins are not highly
conserved (P. aeruginosa and R. capsulatus XdhC proteins
are only 36% identical) and not even present in the xdh
operon (Fig. 1) of some bacteria. Therefore, design of an
optimal probe is problematic.
The sequence of C. acidovorans XDH shows a high level
of identity around the molybdopterin (Fig. 9) and [2Fe)2S]
centers (residues 45–56, 130–140, 170–182) with other
structurally characterized XDHs, despite the lower overall
similarity (43% identity/55% similarity to R. capsulatus
XDH [6] and 31% identity/46% similarity to the bovine
enzyme [5]). An example of this structural identity is shown
in Fig. 9, where the structure around the MPT site in bovine
XDH is shown and the corresponding amino acids in
C. acidovorans are shown in parenthesis. The residues
around the Mo site are identical.

The results from this study show the development of an
XDH expression system that should be beneficial for future
detailed mutagenesis studies on the mechanism of the
C. acidovorans recombinant enzyme with its direct applica-
bility to the mechanism of the eukaryotic enzyme. Addi-
tionally, this system should prove of value as a tool to
explore the function of XdhC in its role of assembly of a
functional MPT active site in the xanthine oxidoreductase
class of enzymes.
Acknowledgements
This study was supported by a grant from the National Institutes of
Health (GM-29433).
References
1. Harrison, R. (2002) Structure and function of xanthine oxido-
reductase: where are we now? Free Radic. Biol. Med. 33, 774–797.
2. Hille, R. (1996) The mononuclear molybdenum enzymes. Chem.
Rev. 96, 2757–2816.
3. Leimkuhler, S., Kern, M., Solomon, P.S., McEwan, A.G.,
Schwarz, G., Mendel, R.R. & Klipp, W. (1998) Xanthine dehy-
drogenase from the phototrophic purple bacterium Rhodo-
bacter capsulatus is more similar to its eukaryotic counterparts
Fig. 9. Schematic representation of the active site of the Comamo-
nas acidovorans XDH Mo center. ThefigureisbasedontheBos taurus
XDH structure (residues of which are shown in parenthesis) with
corresponding conserved C. acidovorans XDH residues (shown in
bold).
Ó FEBS 2003 Influence of xdhC on functional XDH expression (Eur. J. Biochem. 270) 4753
than to prokaryotic molybdenum enzymes. Mol. Microbiol. 27,
853–869.
4. Massey, V. & Edmondson, D. (1970) On the mechanism of

inactivation of xanthine oxidase by cyanide. J. Biol. Chem. 245,
6595–6598.
5. Enroth, C., Eger, B.T., Okamoto, K., Nishino, T. & Pai, E.F.
(2000) Crystal structures of bovine milk xanthine dehydrogenase
and xanthine oxidase: structure-based mechanism of conversion.
Proc.NatlAcad.Sci.USA97, 10723–10728.
6. Truglio, J.J., Theis, K., Leimku
¨
hler, S., Rappa, R., Rajagopalan,
K.V. & Kisker, C. (2002) Crystal structures of the active and
alloxanthine-inhibited forms of xanthine dehydrogenase from
Rhodobacter capsulatus. Structure 10, 115–125.
7. Nishino,T.,Saito,T.,Amaya,Y.,Kawamoto,S.,Ikemi,Y.&
Nishino, T. (1994) Overexpression and characterization of xan-
thine dehydrogenase in a baculovirus-insect cell system. In Flavins
and Flavoproteins 1993. Proceedings of the 11th International
Symposium on Flavins and Flavoproteins (Yagi, K., ed.), pp. 711–
714. de Gruyter, Berlin.
8. Nishino, T., Kashima, Y., Okamoto, K., Iwasaki, T. &
Nishino, T. (1997) The monomeric form of xanthine dehydro-
genase expresses in baculovirus-insect cell system. In Flavins and
Flavoproteins 1996 (Stevenson, K.J., Massey, V. & Williams,
C.H.J., eds), pp. 843–846. University of Calgary Press, Calgary.
9. Iwasaki,T.,Okamoto,K.,Nishino,T.,Mizushima,J.&Hori,H.
(2000) Sequence motif-specific assignment of two [2Fe)2S] clus-
ters in rat xanthine oxidoreductase studied by site-directed muta-
genesis. J. Biochem. (Tokyo) 127, 771–778.
10. Adams, B., Smith, A.T., Doyle, W.A., Bray, R.C., Ryan, M.,
Harrison,R.,Wolstenholme,A.J.,Romao,M.J.,Huber,R.,
Demais, S. & Scazzocchio, C. (1997) Expression of wild-type and

mutated Drosophila melanogaster xanthine dehydrogenases in
Aspergillus nidulans. Biochem. Soc. Trans. 25, 520S.
11. Doyle, W.A., Burke, J.F., Chovnick, A., Dutton, F.L., Russell, C.,
Whittle, J.R. & Bray, R.C. (1996) Engineering and expression in
Drosophila melanogaster of a xanthine dehydrogenase (rosy)
variant. Biochem. Soc. Trans. 24,31S.
12. Doyle, W.A., Burke, J.F., Chovnick, A., Dutton, F.L., Whittle,
J.R. & Bray, R.C. (1996) Properties of xanthine dehydrogenase
variants from rosy mutant strains of Drosophila melanogaster and
their relevance to the enzyme’s structure and mechanism. Eur. J.
Biochem. 239, 782–795.
13. Bilous, P.T. & Weiner, J.H. (1988) Molecular cloning and
expression of the Escherichia coli dimethyl sulfoxide reductase
operon. J. Bacteriol. 170, 1511–1518.
14. Temple, C.A., Graf, T.N. & Rajagopalan, K.V. (2000) Optimi-
zation of expression of human sulfite oxidase and its molybdenum
domain. Arch. Biochem. Biophys. 383, 281–287.
15. Wahl, R.C., Warner, C.K., Finnerty, V. & Rajagopalan, K.V.
(1982) Drosophila melanogaster ma-1 mutants are defective in
the sulfuration of desulfo Mo hydroxylases. J. Biol. Chem. 257,
3958–3962.
16. Amrani, L., Primus, J., Glatigny, A., Arcangeli, L., Scazzocchio,
C. & Finnerty, V. (2000) Comparison of the sequences of the
Aspergillus nidulans hxB and Drosophila melanogaster ma-1 genes
with nifS from Azotobacter vinelandii suggests a mechanism for the
insertion of the terminal sulphur atom in the molybdopterin
cofactor. Mol. Microbiol. 38, 114–125.
17. Ichida, K., Matsumura, T., Sakuma, R., Hosoya, T. & Nishino, T.
(2001) Mutation of human molybdenum cofactor sulfurase gene is
responsible for classical xanthinuria type II. Biochem. Biophys.

Res. Commun. 282, 1194–1200.
18. Watanabe, T., Ihara, N., Itoh, T., Fujita, T. & Sugimoto, Y.
(2000) Deletion mutation in Drosophila ma-1 homologous, puta-
tive molybdopterin cofactor sulfurase gene is associated with
bovine xanthinuria type II. J. Biol. Chem. 275, 21789–21792.
19. Bittner, F., Oreb, M. & Mendel, R.R. (2001) ABA3 is a
molybdenum cofactor sulfurase required for activation of
aldehyde oxidase and xanthine dehydrogenase in Arabidopsis
thaliana. J. Biol. Chem. 276, 40381–40384.
20. Muhlenhoff, U. & Lill, R. (2000) Biogenesis of iron-sulfur proteins
in eukaryotes: a novel task of mitochondria that is inherited from
bacteria. Biochim. Biophys. Acta 1459, 370–382.
21. Leimkuhler, S. & Klipp, W. (1999) Role of XDHC in molybde-
num cofactor insertion into xanthine dehydrogenase of Rhodo-
bacter capsulatus. J. Bacteriol. 181, 2745–2751.
22. Xiang, Q. & Edmondson, D.E. (1996) Purification and char-
acterization of a prokaryotic xanthine dehydrogenase from
Comamonas acidovorans. Biochemistry 35, 5441–5450.
23. Leimkuhler, S., Hodson, R., George, G.N. & Rajagopalan, K.V.
(2003) Recombinant Rhodobacter capsulatus xanthine dehydro-
genase, a useful model system for the characterization of protein
variants leading to xanthinuria I in humans. J. Biol. Chem. 278,
20802–20811.
24. Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989) Molecular
Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, New York.
25. Laemmli, U.K. (1970) Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227, 680–685.
26. Xiang, Q. (1996) Comamonas acidovorans Xanthine Dehydro-
genase: Phosphorylation, Catalytic, and Structural Studies.Emory

University, Atlanta.
27. Hecker, K.H. & Roux, K.H. (1996) High and low annealing
temperatures increase both specificity and yield in touchdown and
stepdown PCR. Biotechniques 20, 478–485.
28. Gornall, A.G., Bardawill, C.J. & David, M.M. (1949) Determi-
nation of serum proteins by means of the biuret reaction. J. Biol.
Chem. 177, 751–766.
29. Bearden, J.C. Jr (1978) Quantitation of submicrogram quantities
of protein by an improved protein-dye binding assay. Biochim.
Biophys. Acta 533, 525–529.
30. Cardenas, J. & Mortenson, L.E. (1974) Determination of moly-
bdenum and tungsten in biological materials. Anal. Biochem. 60,
372–381.
31. Vanoni, M.A., Edmondson, D.E., Zanetti, G. & Curti, B. (1992)
Characterization of the flavins and the iron-sulfur centers of glu-
tamate synthase from Azospirillum brasilense by absorption,
circular dichroism, and electron paramagnetic resonance spec-
troscopies. Biochemistry 31, 4613–4623.
32. Vanoni, M.A., Verzotti, E., Zanetti, G. & Curti, B. (1996) Prop-
erties of the recombinant beta subunit of glutamate synthase. Eur.
J. Biochem. 236, 937–946.
33. Komai, H., Massey, V. & Palmer, G. (1969) The preparation and
properties of deflavo xanthine oxidase. J. Biol. Chem. 244,
1692–1700.
34. Hille, R., Hagen, W.R. & Dunham, W.R. (1985) Spectroscopic
studies on the iron-sulfur centers of milk xanthine oxidase. J. Biol.
Chem. 260, 10569–10575.
35. Palmer, T., Vasishta, A., Whitty, P.W. & Boxer, D.H. (1994)
Isolation of protein FA, a product of the mob locus required for
molybdenum cofactor biosynthesis in Escherichia coli. Eur. J.

Biochem. 222, 687–692.
36. Garrett,R.M.&Rajagopalan,K.V.(1994)Molecular cloning of rat
liver sulfite oxidase. Expression of a eukaryotic Mo-pterin-con-
taining enzyme in Escherichia coli. J. Biol. Chem. 269, 272–276.
37. Gardlik, S. & Rajagopalan, K.V. (1990) The state of reduction of
molybdopterin in xanthine oxidase and sulfite oxidase. J. Biol.
Chem. 265, 13047–13054.
38. Berglund,L.,Rasmussen,J.T.,Andersen,M.D.,Rasmussen,M.S. &
Petersen, T.E. (1996) Purification of the bovine xanthine
oxidoreductase from milk fat globule membranes and cloning of
complementary deoxyribonucleic acid. J. Dairy Sci. 79, 198–204.
4754 N. V. Ivanov et al. (Eur. J. Biochem. 270) Ó FEBS 2003