(Received for publication, January 24, 1997, and in revised form, February 19, 1997)
From the Medical Nobel Institute for Biochemistry,
Department of Medical Biochemistry and Biophysics, Karolinska
Institutet, S-171 77 Stockholm, Sweden and the ¶ Laboratory of
Chemistry I, Department of Medical Biochemistry and Biophysics,
Karolinska Institutet, S-171 77 Stockholm, Sweden
Glutaredoxin 2 (Grx2) from
Escherichia coli catalyzes GSH-disulfide oxidoreductions
via two redox-active cysteine residues, but in contrast to glutaredoxin
1 (Grx1) and glutaredoxin 3 (Grx3), is not a hydrogen donor for
ribonucleotide reductase. To characterize Grx2, a chromosomal fragment
containing the E. coli Grx2 gene (grxB) was
cloned and sequenced. grxB (645 base pairs) is located between the rimJ and pyrC genes while an open
reading frame immediately upstream grxB encodes a novel
transmembrane protein of 402 amino acids potentially belonging to class
II of substrate export transporters. The deduced amino acid sequence
for Grx2 comprises 215 residues with a molecular mass of 24.3 kDa.
There is almost no similarity between the amino acid sequence of Grx2
and Grx1 or Grx3 (both 9-kDa proteins) with the exception of the active
site which is identical in all three glutaredoxins
(C9PYC12 for Grx2). Only limited similarities
were noted to glutathione S-transferases (Grx2 amino acids
16-72), and protein disulfide isomerases from different organisms
(Grx2 amino acids 70-180). Grx2 was overexpressed and purified to
homogeneity and its activity was compared with those of Grx1 and Grx3
using GSH, NADPH, and glutathione reductase in the reduction of 0.7 mM -hydroxyethyl disulfide. The three glutaredoxins had
similar apparent Km values for GSH (2-3
mM) but Grx2 had the highest apparent
kcat (554 s
1). Expression of two
truncated forms of Grx2 (1-114 and 1-133) which have predicted
secondary structures similar to Grx1 (
) gave rise
to inclusion bodies. The mutant proteins were resolubilized and
purified but lacked GSH-disulfide oxidoreductase activity. The latter
should therefore require the participation of amino acid residues from
the COOH-terminal half of the molecule and is probably not confined to
a Grx1-like NH2-terminal subdomain. Grx2 being radically
different from the presently known glutaredoxins in terms of molecular
weight, amino acid sequence, catalytic activity, and lack of a
consensus GSH-binding site is the first member of a novel class of
glutaredoxins.
Glutaredoxin (Grx)1 was discovered as
a glutathione-dependent hydrogen donor for
Escherichia coli ribonucleotide reductase (1-3). The first
isolated glutaredoxin (glutaredoxin 1) (Grx1), is a 9-kDa protein with
two catalytic cysteine residues in the sequence Cys-Pro-Tyr-Cys (4).
Apart from its protein disulfide reductase activity with ribonucleotide
reductase, Grx1 is also a general GSH-disulfide oxidoreductase,
reducing disulfides like -hydroxyethyl disulfide (HED) in a coupled
system with GSH, NADPH, and glutathione reductase (3, 5) (HED assay,
Equations 1 and 2),
![]() |
(Eq. 1) |
![]() |
(Eq. 2) |
In a crude extract of wild-type E. coli B, Grx1 constitutes only about 2% of the total GSH-disulfide oxidoreductase activity (HED assay) (5, 6). Later work resulted in the isolation of two additional glutaredoxins which accounted for the major general GSH-disulfide oxidoreductase activity (6). From their order of elution on Sephadex G-50 the new glutaredoxins were called glutaredoxin 2 (Grx2) and glutaredoxin 3 (Grx3). NH2-terminal sequencing of Grx2 and Grx3 demonstrated that both enzymes contained the typical glutaredoxin motif (Cys-Pro-Tyr-Cys), although Grx2 had an atypical size (Mr 27,000) for glutaredoxin. Grx3 (9 kDa) with 82 residues showed 33% sequence identity and a similar secondary structure and tertiary fold (7) as the well characterized Grx1 (8, 9). Thus both Grx1 and Grx3 have the thioredoxin fold (10, 11) and are members of the thioredoxin (Trx) superfamily of disulfide oxidoreductases. In contrast to Grx3, which has about 5% of the catalytic activity of Grx1 as a disulfide reductant of ribonucleotide reductase, Grx2 lacks such activity (6). The relative abundance of Grx2 in E. coli (6) makes further investigations of the structure and function of this unknown protein necessary to understand the SH-metabolism in the cell. In this study we have cloned and sequenced the gene coding for Grx2 (grxB) and another gene located immediately upstream grxB coding for a potential transmembrane protein. Grx2 showed large differences to previously known glutaredoxins.
-Hydroxyethyl disulfide (HED) was from Aldrich. NADPH and
glutathione reductase (yeast) were from Sigma. Diaflo YM10 membranes were from Amicon. E. coli Grx1 and Grx3, Trx, and
thioredoxin reductase were purified to homogeneity as described (12).
Oligonucleotides were synthesized by Pharmacia Biotech Ltd. Restriction
enzymes, T4 DNA ligase, and Taq polymerase were from
Promega.
XL-1 blue and DH5 were
used for plasmid propagation and cloning purposes. Strain GI698 (13)
was a gift Dr. J. McCoy (Genetics Institute, Boston, MA) and was used
for overexpression of Grx2.
Total protein was
measured using the Bradford protein assay (14) adapted for use on
microtiter plates. For pure Grx2, a molar extinction coefficient of
21,860 M1·cm
1 at 280 nm was
used.
A nearly homogenous preparation (more than 90% pure as judged by SDS-PAGE) of E. coli Grx2 (5 nmol in 50 µl) purified as described (6), was reduced in 0.5 ml of 50 mM Tris-Cl, pH 8.0, by incubation for 4 h at 4 °C in the presence of 1 mM dithiothreitol. Carboxymethylation was performed by the addition of 300 µl of 6 M guanidine hydrochloride, 0.4 M Tris-Cl, pH 8.1, 2 mM EDTA, 5 mM neutralized 14C-iodoacetic acid (Amersham, approximately 2400 cpm/nmol), and incubation for 4 h at 4 °C. Reagents and contaminating proteins were removed by reverse-phase chromatography on a C4 (Vydac) HPLC column equilibrated in 0.1% trifluoroacetic acid in water and eluted with a linear gradient of acetonitrile (0-40% during 60 min, 1 ml/min) containing 0.1% trifluoroacetic acid.
Peptide Generation, Purification, Nomenclature, and Sequence AnalysisHPLC-purified 14C-carboxymethylated Grx2 was
digested for 4 h at 37 °C with Lys-C endoprotease (Wako), at an
enzyme to protein ratio of 1:10 in 0.1 M ammonium
bicarbonate, pH 8.0, containing 1 M guanidine
hydrochloride. The resulting peptides (K peptides, Fig. 1) were
purified by reverse-phase (C18) HPLC using conditions described above.
Peptides (D peptides, Fig. 1) were generated also by cleavage with
Asp-N endoprotease (Boehringer-Mannheim) and purified on a C4 HPLC
column, see above. Amino acid sequence analysis of all peptides was
performed using an Applied Biosystems 470A sequencer and
phenylthiohydration detection was by reverse-phase HPLC in a
Hewlett-Packard 1090 system (15).
Construction of PCR Primers and Cloning of an 180-Base Pair PCR Product for Grx2
Two forward primers corresponding to the
published NH2-terminal sequence of Grx2 (6) were
synthesized and named N1 (amino acids 1-8:
5-ATGAAACTGTACATCTA(C/T)GA(C/T)CA-3
) and N2 (amino acids 10-15:
5
-CCGTACTGCCTGAA(A/G)GC-3
). The reverse primers were based on the
peptide K8 (DDSRYMPESMDIVHYVDK) and were named C1 (amino acids 4-9 of
K8: 5
-CATAGATTCCGGCAT(G/A)TA-3
) and C2 (amino acids 9-16 of K8:
5
-TCAACGTAGTGAACGAT(G/A)TCCAT-3
). A 28-cycle PCR reaction (94 °C
for 1 min, 46 °C for 1.2 min, 72 °C for 2 min) was run against a
preparation of wild type (strain K12) DNA using all four combinations
of primers. Amplified DNA using the N2/C1 and N2/C1 primers was
subcloned into pGEM-T vector (Promega) and sequenced.
An 180-bp PCR product resulting from amplification using the
N2 and C1 primers contained sequences corresponding to peptides from
Grx2. The PCR product was labeled with 32P by random
priming and used as a probe to screen a nitrocellulose filter
containing DNA from the Kohara phage library (16). Phage clone 233 was
selected and propagated in E. coli strain Y1090. Phage
DNA was purified using the plate lysate method (17), with the
modification of resuspending pelleted phage particles in
diluent
(instead of TE) after DNase I treatment. The modification resulted to
higher yields of non-degraded phage DNA.
Phage DNA from clone 233 was digested with BamHI and EcoRI, separated on a 1% agarose gel, transferred to a nitrocellulose membrane, and finally probed using the 180-bp PCR fragment (N2 and C1 primers). A ~4.2-kilobase band which hybridized strongly with the probe was cloned into vector pGEM-3Z (Promega) and used for sequence analysis. The nucleotide sequence of the cloned fragment was determined using the Taq DyeDeoxy terminator cycle sequencing on an automated laser fluorescent DNA sequencer (A.L.F., Pharmacia) according to the protocol supplied by the manufacturer. Primers (fluorescent) were made initially based on the 180-bp PCR fragment, and later on the sequences revealed upstream and downstream. Sequencing stopped after the known pyrC and rimJ regions were reached.
Overexpression of Grx2grxB was amplified using
primers containing NdeI (G2-FNdeI primer:
5-TGGAGGAGTCATATGAAGCTATAC-3
) and BamHI (G2-RCBamHI
primer: 5
-CGCGGCGGGGGATCCTTAAATCGC-3
) sites at the 5
and 3
termini of the gene, respectively. The amplified fragment was then digested with NdeI and BamHI and was cloned into the
pTRXFUS (13) vector replacing the Trx gene and giving rise the Grx2
expression vector named pe1Grx2. Strain GI698 transformed with pe1Grx2
was grown in 1-liter cultures at 30 °C. Grx2 expression was induced
at an optical density of cells at 600 nm (A600)
between 0.4 and 0.5 with a final concentration of 100 µg/ml
DL-tryptophan. Cells were then grown for up to 16 h in
an orbital shaker at 100 rpm and harvested by centrifugation at
6000 × g for 20 min.
Truncated forms of grxB were constructed using the G2-FNdeI
primer and primers G2del1RC (for 1-114,
5-GAAATATTTGCGGGATCCTTATTAAGAAAACTCATC-3
) or G2del2RC (for 1-133,
5
-AGAGTGGGCCAGGGATCCTTAAAAATTACCCGCGC-3
). Both primers contained a
BamHI site. The expression strategy for the truncated forms
was identical to that used for the wild-type protein with the exception
that temperatures of 20 and 25 °C also were used in separate
experiments for growth and expression.
Cells, 6.2 g, were resuspended in 50 ml of 50 mM Tris-Cl, pH 8.0, and placed on ice. Lysozyme (0.5 mg/ml final), phenylmethylsulfonyl fluoride (1 mM final), and DNase I (50 µg/ml final) were added sequentially after 5 min each and the mixture was left on ice for 1 h. EDTA was added to a final concentration of 1 mM and the cell lysate was thoroughly sonicated (on ice three times at full power for 1 min). The supernatant was collected after centrifugation at 24,000 × g for 35 min at 4 °C and the pellet was washed with 10 ml of 100 mM Tris-Cl, pH 8.0, recentrifuged, and the new supernatant was pooled with the first. The pool was dialyzed overnight against 20 mM Tris base (pH above 10) and was applied to a 200-ml DE52 column (Whatman) equilibrated with the same buffer. Grx2 bound to the column and was eluted by 50 mM Tris-Cl, pH 8.0, in a volume of approximately 2 liters which after concentration through a tangential flow cartridge with a cut-off of 10 kDa (Prep/Scale-TFF cartridge, Millipore) was applied to a column (100 cm × 1.6 cm2) of Sephadex G-50 superfine (Pharmacia) with 100 mM potassium phosphate, pH 7.0, as equilibration and running buffer.
Preparation of Soluble Truncated Forms of Grx2The 1-114 and 1-133 truncated forms of Grx2 were recovered from the insoluble fraction after cell lysis. The insoluble pellet from a 1-liter culture was washed in 20 ml of 100 mM Tris-Cl, pH 8.0, 1 mM EDTA, recentrifuged, and solubilized with 10 ml of 100 mM Tris-Cl, pH 8.0, 1 mM EDTA, 100 mM dithiothreitol, 8 M urea. The urea extract was diluted 100-fold by dropwise addition to 1 liter of 100 mM Tris-Cl, pH 8.0, 1 mM EDTA at 4 °C. The mixture was concentrated to 10 ml on a Diaflo membrane and centrifuged at 24,000 × g for 35 min at 4 °C. The supernatant contained the Grx2 1-114 or Grx2 1-133 fragment in soluble form. The purity of these preparations was more than 50% as judged by SDS-PAGE.
Determination of Glutaredoxin Activity (HED assay)A fresh
mixture of 1 mM GSH, 0.2 mM NADPH, 2 mM EDTA, 0.1 mg/ml bovine serum albumin, and 6 µg/ml
yeast GR was prepared in 100 mM Tris-Cl, 2 mM
EDTA, pH 8.0 (3). To 500 µl of this mixture in semimicro cuvettes,
HED was added to a final concentration of 0.7 mM. After 3 min glutaredoxin was added to the sample cuvettes and the decrease in
A340 was recorded the following 5 min using a
Zeiss PMQ3 spectrophotometer. Activity was expressed as micromoles of
NADPH oxidized per min using a molar extinction coefficient of 6200 M1·cm
1. Steady-state kinetics
measurements were performed in a final volume of 100 µl using a
Molecular Devices Thermomax microplate reader. Values of
A340 were multiplied by a factor of 4.3 to give the
A340 of a cuvette with a path length
of 1 cm. All activity measurements were made at 25 °C.
This was done
spectrophotometrically with 1 mM
5,5-dithiobis(2-nitrobenzoic acid) in 200 mM Tris-Cl, pH
8.0, 6 M guanidine hydrochloride, using a molar extinction
coefficient of 13,600 M
1·cm
1
at 412 nm.
The ability of Grx2 to serve
as a substrate of thioredoxin reductase was determined in the reduction
of 5,5-dithiobis(2-nitrobenzoic acid). This was followed at 412 nm in
100 mM Tris-Cl, pH 8.0, 2 mM EDTA, 0.1 mg/ml
bovine serum albumin, 0.2 mM NADPH, 35 nM E. coli thioredoxin reductase, and 0.5 mM
5,5
-dithiobis(2-nitrobenzoic acid) as described (12). E. coli thioredoxin was used as a positive control.
Measurements were done at 340 nm using the HED assay for glutaredoxin (100 mM Tris-Cl, pH 8.0, 2 mM EDTA, 0.1 mg/ml bovine serum albumin, 0.2 mM NADPH, 1 mM GSH, 6 µg/ml glutathione reductase) (3) but with bovine insulin (33 µM) in place of HED as a disulfide substrate.
Preparation of Cellular Fractions from E. coliPeriplasmic fractions were prepared after three cycles of quick freezing and thawing in dry ice-ethanol bath, in phosphate-buffered saline at an A600 of 40 (18). The supernatant after centrifugation was collected as the periplasmic lysate fraction. To prepare osmotic shock fractions, cells were initially resuspended (A600 = 40) in ice-cold 20% sucrose, 20 mM Tris-Cl, pH 8.0, with or without 1 mM EDTA for 10 min. They were then resuspended in ice-cold 100 mM Tris-Cl, pH 8.0, 1 mM EDTA and left on ice for 10 min. Cells were finally centrifuged and the supernatant was collected as the osmotic shock fraction (13).
SDS-PAGEThis was performed according to Laemmli (19) using the Bio-Rad mini gel apparatus. Gels were stained with Coomassie Blue.
The sequence for the 18 NH2-terminal residues of Grx2 including its active site was
determined previously (6). We now digested Grx2 with either Lys-C
endoprotease or Asp-N endoprotease and separated the peptides by HPLC.
Altogether 26 peptides were analyzed by automated Edman degradation
(Fig. 1). Based on the NH2-terminal sequence
and internal peptide K8, primers were designed (N2 and C1) which
following PCR resulted in the generation of an 180-bp PCR product
containing part of the sequence of grxB. The 180-bp PCR
product was used as a probe to screen the Kohara phage library (16) and
a positive clone ( 233) was identified. After cleavage of
DNA
with EcoRI and BamHI a 4.2-kilobase fragment
which hybridized with the PCR probe was cloned and sequenced. The novel
sequence revealed in the 4.2-kilobase fragment was named
grxB region (1943 bp) (Fig. 1) and contained the sequences
of grxB and another open reading frame (ORF) upstream
grxB.
The 645-bp ORF for grxB encodes a 215-amino acid residue
protein with a predicted molecular mass of 24.3 kDa. The deduced amino
acid sequence was in agreement with the available amino acid sequence
data from the peptides derived from purified Grx2. The
NH2-terminal methionine is coded for by GTG. A consensus
ribosomal binding site (GGAGG) is located 6 bp upstream of the
initiator GTG codon. There are two 35 TTGACa promoter consensus
sequences upstream of grxB there but no obvious
10 TAtAaT
consensus (Fig. 1).
grxB is partially overlapping with the pyrC region data base
entry (Fig. 2) and is located 802 bp upstream from the
actual E. coli gene for dihydroorotase (pyrC)
(23.4 min) (20). The gene encoding the NH2-terminal
acetylase of ribosomal protein S5 (rimJ) (23.7 min) (21)
starts 1216 bp upstream from grxB, but is translated from
the complementary DNA strand. In fact both grxB and
pyrC regions are translated anticlockwise.
Sequence and Structure of a Novel Putative Membrane Protein
An ORF corresponding to 402 amino acids is located
immediately upstream of grxB, before rimJ (Figs.
1 and 2). A GGAGG consensus ribosomal binding site is located three
bases upstream from the first Met residue or 12 bp upstream to an
alternative GTG-encoded Met. The TMAP program for the prediction of
transmembrane proteins (22) aligned the putative protein to well
established transmembrane proteins and a membrane orientation model was
constructed (Fig. 3). The grxB region
containing grxB and the preceding ORF has been submitted to
the EMBL data base with accession number X92076[GenBank].
Expression and Purification of Recombinant Grx2
The gene
sequence used for the overexpression of Grx2 was identical to the
chromosomal one with the exception that the GTG initiator codon was
substituted with ATG to increase expression levels since GTG has been
reported as a less efficient start codon compared with ATG (23). Using
the overexpression system for Trx fusions (13) with Grx2 replacing Trx,
recombinant Grx2 was expressed in a soluble state as high as 50% of
total cell protein (Fig. 4). Chromatographies on columns
of DE52 cellulose and Sephadex G-50 superfine resulted in a homogeneous
protein. A yield of 40% from the purification procedure was obtained.
A summary of the purification procedure and yields is shown in Table
I. The molecular mass of the recombinant Grx2 was
determined as 24.3 kDa by matrix-assisted laser desorption/ionization
mass spectrometry. This value is identical to that predicted from the
genomic sequence. Peptide sequencing showed that the initiation Met was
present in preparations of the recombinant enzyme as was also the case
for protein isolated from E. coli cells (6). The specific
activity of the recombinant Grx2 (316 units/mg, Table I) is identical
to the value published before (6) for the wild type enzyme which was
based on a Mr of 27,000 rather than the correct
value of 24,300.
|
Previous
results (6) suggested that a large part of the GSH-disulfide
oxidoreductase activity in E. coli cells measured by the
standard HED assay (Equations 1 and 2) is due to Grx2. To compare the
steady-state contribution of HED reducing activities of the three
E. coli glutaredoxins, we determined their apparent Km and kcat values for GSH
using the standard HED assay which employs final concentrations of 1 mM GSH and 0.7 mM HED. Comparison was also made
with human glutaredoxin (hGrx) (24, 25). Due to the spontaneous
reaction between HED and GSH leading to increased NADPH consumption, we
only determined rates up to 4.5 mM GSH (Fig.
5). The kinetic data (Fig. 5, Table II)
showed that Grx2 had the lowest apparent Km for GSH
and the highest turnover.
|
Sequence alignment of Grx2 with Grx1 and Grx3 showed
surprisingly very little homology and identities (Fig.
6). The active site is identical in all three proteins
and since Grx2 contains two Cys residues, these must be the catalytic
ones located in a similar relative position as in Grx1 and Grx3.
Sequence analysis using data base programs revealed some limited
homologies to glutathione S-transferases (GSTs) and protein
disulfide isomerases (PDIs) (Fig. 6). Since Grx2 seemed to contain
Grx-like and PDI-like parts (Fig. 6), we constructed and expressed
relevant truncated mutant proteins to examine their possible functional
autonomy. The first (amino acids 1-114) mutant corresponded to a part
of Grx2 which is long enough to include a Grx-like domain (~80 amino
acids). The second mutant was extended to amino acid 133 to include the part with homology to bovine PDI. Both mutants were expressed at high
levels but in insoluble states (Fig. 4). Resolubilization from 8 M urea gave 50% pure preparations but no activity could be
detected with the standard HED assay for concentrations of the
truncated forms up to 10 µM (assuming a 50% purity, data
not shown). In comparison, activity of the least active of the four glutaredoxins, Grx1, is detected readily at concentrations of 0.001 µM.
Grx2 as a Reductant of Insulin Disulfides
Trx and Grx1 reduce
disulfides in insulin, with Grx1 being a much less efficient enzyme
(3). Our results show that Grx2 and Grx3 did not reduce insulin using 1 mM GSH (Fig. 7). This probably reflects an
unfavorable enzyme substrate interaction and/or an elevated redox
potential for Grx2 and Grx3. The fact that purified Grx2 was obtained
in a fully reduced state, as measurements of free SH-groups showed,
suggests that Grx2 has a high redox potential and is rather
oxidizing.
Grx2 Is Not a Substrate for Thioredoxin Reductase
Recombinant
alkaline phosphatase can be expressed as an active protein with all its
correct disulfides in E. coli trxB mutants, but this cannot
happen in trxA mutants lacking thioredoxin (26). This
suggests the presence of an alternative substrate for thioredoxin reductase which reduces and inactivates alkaline phosphatase (26). Grx2
did not reduce 5,5-dithiobis(2-nitrobenzoic acid), demonstrating that
it was not a substrate for thioredoxin reductase (data not shown). To
examine whether GSH was required for thioredoxin reductase-Grx2 interactions, standard HED assays were carried with thioredoxin reductase substituting for glutathione reductase. Grx2 was not reduced
under these conditions as measured at 340 nm showing that it did not
react with thioredoxin reductase through GSH.
To examine the
intracellular localization of Grx2, periplasmic and cytosolic fractions
of non-transformed E. coli cells and transformants were
analyzed by SDS-PAGE (Fig. 8). Grx2 was clearly cytosolic. The enzyme was not released (even when overexpressed) by
osmotic shock conditions which is different from Trx (1, 13) and Grx1
(1).
This paper describes the cloning and sequence determination of a genomic region containing grxB and presents findings on the characterization of recombinant Grx2. Previously two other glutaredoxins from E. coli, both 9-kDa proteins, were shown to have the same active site sequence Cys-Pro-Tyr-Cys and to catalyze GSH-disulfide oxidoreductions via a GSH-binding site (7, 27). Grx2 contains the same active site sequence Cys-Pro-Tyr-Cys as the other two glutaredoxins and is highly active in the general GSH-disulfide oxidoreductase assay with the small disulfide HED as a substrate. Grx2, however, has a larger size and lacks activity as a hydrogen donor (protein disulfide reductase) for ribonucleotide reductase (6). Alignment of Grx2 with other proteins showed no further extensive homology to known glutaredoxins or thioredoxins (Fig. 6). Grx2 therefore lacks the conserved sequences of Grx1 and Grx3 defining a GSH-binding site previously experimentally verified by NMR for the mixed disulfide between E. coli Grx1 C14S and GSH (28). Amino acid residues of Grx2 that may be the homologues to amino acids of Grx1 involved in binding to GSH (underlined) are Y11 (13 in Grx1), QVP49 (P60 in Grx1), and YVD70 (74 in Grx1).
Data base searches showed that Grx2 was weakly homologous to sequences
of two GSH S-transferases (GSTs) (Grx2 amino acids 47-69
and 86-112), a relative of GST, crystallin SL-11 (28) (Grx2 amino
acids 16-71), and PDIs of different organisms (70-180) (Fig. 6). The
significance of this is unknown. Secondary structure prediction using
the method of Rost and Sander (29) suggested that the first third of
Grx2 (2-70) contained an alternation of
-sheets and
-helices
(
), while the remaining part (84-212) was
predominantly helical (Fig. 6). The
prediction for the NH2-terminal third of the molecule is similar to the
characteristic
secondary structure of the
thioredoxin fold found in E. coli Trx (10), Grx1 (8), Grx3
(7), glutathione peroxidase (30), and GSH S-transferases
(GST) (31). In the latter a first Trx-like domain (~80 amino acids)
is connected (through a short linker of 8 amino acids for GSTµ) (32)
to a second domain (~130 amino acids) composed of
helices (31).
Although the amino acid sequence of Grx2 was markedly different from
other members of the Trx superfamily or GSTs including the known
E. coli GST (Fig. 6) (33), the NH2-terminal
third of the molecule may have the thioredoxin fold. E. coli
DsbA, for example, has almost no sequence homology to Trx (34) but
contains a Trx-like domain. To examine whether Grx2 contained a fully
active Trx/Grx-like domain at its NH2 terminus, we
overexpressed two truncated mutant proteins. The first (1-114)
included the putative
fold and the second
(1-133) included a part with homology to bovine PDI (Fig. 6). Both
overexpressed protein fragments were inactive in the glutaredoxin assay
(HED assay). These data suggested that the GSH-disulfide oxidoreductase
activity of Grx2 requires parts of the COOH-terminal half of the
molecule and is not confined to a Grx1-like NH2-terminal
subdomain. In comparison, human PDI contained distinct Trx domains
which can be expressed as soluble proteins which catalyze
thiol-disulfide exchange reactions (35). Grx2 therefore appears more
similar to GSTs in which the GSH-binding site is not confined
exclusively to the
Grx-like first domain but also
includes residues from the second domain (
4/
5 helix-turn-helix
segment and COOH terminus) (31, 32). Grx2 residues that may be the
homologues of GST residues known to interact with GSH (underlined) are
Y6 (6 in GSTµ, 7 in GST
, and
8 in GST
) (31), V in QVP49
(QP55 in seven different GST
s) (31), and
perhaps the W in WL90 (L47 in
eight different GSTµs) (31). Biochemically Grx2 was different from
GSTs as it was inactive in the standard 1-chloro-2,4-dinitrobenzene
assay for GST activity (36) and did not form the
-complex with
1,3,5-trinitrobenzene (37) as GSTs do.2
Furthermore, Grx2 is a monomeric and not a dimeric enzyme either in its
reduced or oxidized form as determined by Superdex G-75 chromatography
in either 20 or 100 mM potassium phosphate, pH 7.0. In
comparison, all known GSTs form homodimeric complexes in solution
(31).
The ORF immediately upstream grxB was predicted as a
transmembrane protein, using the TMAP program for the prediction and alignment of transmembrane proteins (22). Homologies were observed with
other transmembrane transporter proteins and a membrane orientation model was constructed (Fig. 3) after sequence alignment with the metal-tetracycline/H+ antiporter 1 which is thought to
contain 12 membrane spanning segments (38). In this model the ORF
follows the positive inside rule for transmembrane proteins (39) as it
contains one Arg in the periplasm compared with 25 Arg, and Lys in the
cytosol. Similarities with other transmembrane proteins reside on the
sequence AIAFA
between potential 2 and
3
(GXXXD(R/K)XGR(R/K) is the consensus for
transmembrane transporters) (40). Transporters which aligned with the
ORF include four tetracycline antiporters (TetA, B, C, and D), the
quinolone resistance protein from Staphylococcus aureus
(NorA) (41), and the multidrug resistance protein from Bacillus
subtilis (Bmr) (42). Since these transporters belong to class II
of transmembrane proteins (40), we suggest that the ORF before Grx2
also is a member of this class.
The ORF transporter contains 5 Cys residues (Fig. 3), although none
with the thioredoxin CXXC active-site motif. Cysteine residues also were present in other transmembrane proteins that aligned
with the ORF (not shown). Of particular interest is whether the ORF
transporter (63 bp upstream grxB), Grx2, and the protein encoded by the ORF before pyrC (169 bp downstream
grxB) are functionally coupled (in principle the transporter
could export a product delivered from Grx2) and belong to the same
operon. The latter could be possible because although there are two
35 TTGACa promoter consensus sequences upstream from grxB
there is no obvious
10 TAtAaT consensus. Also, there is no obvious
consensus for the termination of transcription downstream from the ORF
before grxB or grxB itself. However, there is a
transcriptional terminator downstream from the ORF before pyrC and expression of the pyrC genes is known to
be independent from genes upstream (20).
The catalytic activity of recombinant Grx2 and three other classical glutaredoxins was compared in the GSH-disulfide oxidoreductase assay varying the concentrations of GSH with the HED concentration constant (0.7 mM). All the different glutaredoxins had similar apparent Km values for GSH (Fig. 5 and Table II) but Grx2 showed the highest apparent kcat. Grx2 previously was calculated to represent a major part of the total glutaredoxin activity in an E. coli Grx1 null mutant (6). Since the physiological concentrations of GSH in E. coli is from 3.5 to 6.6 mM (43), Grx2 should have the potential of being a major catalyst in GSH-utilizing reactions. The situation may be more complicated. Grx1 has been reported (6) to amount to only 2% of total E. coli oxidoreductase activity (HED assay) (5) but its levels are highly variable and depend on other factors (for example, the levels of ribonucleotide reductase, Trx and GSH) which regulate the intracellular redox environment (44). If Grx2 levels are regulated is unknown and requires further studies.
The biological function of Grx2 is unknown and will require further studies. If Grx2 has a high redox potential it may be involved in GSH-dependent thiol-disulfide exchange reactions making disulfides in proteins. Such reactions and other potential physiological roles linked to the transporter can be addressed by gene inactivation experiments currently in progress.
We are grateful to Carina Jonsson for excellent technical assistance, Valentina Bonetto for the determination of the molecular mass of recombinant Grx2 by matrix-assisted laser desorption/ionization mass spectrometry, and Dr. Ralf Morgenstern for the TNB and CDNB assays. The generous help of Malin Rohdin is also acknowledged.