Identification, expression pattern, and characterization of mouse glutaredoxin 2 isoforms.

Abstract

Glutaredoxin 2 (Grx2) is a glutathione-dependent oxidoreductase involved in the maintenance of mitochondrial redox homeostasis. Grx2 was first characterized as mitochondrial protein, but alternative mRNA variants lacking the transit peptide-encoding first exon were demonstrated for human and proposed for mouse. We systematically screened for alternative transcript variants of mouse Grx2. We identified a total of six exons, three constitutive (II, III, and IV), two alternative first exons (exons Ia and Ic), and one single-cassette exon (exon IIIb) located between exons III and IV. Exons Ic and IIIb are not present in the human genome; mice lack human exon Ib. The six exons give rise to five transcript variants that encode three protein isoforms: mitochondrial Grx2a, a cytosolic isoform that is homologous to the cytosolic/nuclear human Grx2c and present in specific cells of many tissues and the testis-specific isoform Grx2d that is unique to mice. Mouse Grx2c can form an iron/sulfur cluster-bridged dimer, is enzymatically active as a monomer, and can donate electrons to ribonucleotide reductase. Testicular cells lack mitochondrial Grx2a but contain cytosolic Grx2. Prominent immunostaining was detected in spermatogonia and spermatids. These results provide evidence for additional functions of Grx2 in the cytosol, in cell proliferation, and in cellular differentiation.

Full text

Page 1

ANTIOXIDANTS & REDOX SIGNALING
Volume 11, Number 1, 2009
© Mary Ann Liebert, Inc.
DOI: 10.1089/ars.2008.2068
Original Research Communication
Identification, Expression Pattern, and Characterization of
Mouse Glutaredoxin 2 Isoforms
Christoph Hudemann,1Maria Elisabet Lönn,1José Rodrigo Godoy,2Farnaz Zahedi Avval,1
Francisco Capani,3Arne Holmgren,1and Christopher Horst Lillig1,2
Abstract
Glutaredoxin 2 (Grx2) is a glutathione-dependent oxidoreductase involved in the maintenance of mitochon-
drial redox homeostasis. Grx2 was first characterized as mitochondrial protein, but alternative mRNA variants
lacking the transit peptide-encoding first exon were demonstrated for human and proposed for mouse. We sys-
tematically screened for alternative transcript variants of mouse Grx2. We identified a total of six exons, three
constitutive (II, III, and IV), two alternative first exons (exons Ia and Ic), and one single-cassette exon (exon IIIb)
located between exons III and IV. Exons Ic and IIIb are not present in the human genome; mice lack human
exon Ib. The six exons give rise to five transcript variants that encode three protein isoforms: mitochondrial
Grx2a, a cytosolic isoform that is homologous to the cytosolic/nuclear human Grx2c and present in specific
cells of many tissues and the testis-specific isoform Grx2d that is unique to mice. Mouse Grx2c can form an
iron/sulfur cluster–bridged dimer, is enzymatically active as a monomer, and can donate electrons to ribonu-
cleotide reductase. Testicular cells lack mitochondrial Grx2a but contain cytosolic Grx2. Prominent immuno-
staining was detected in spermatogonia and spermatids. These results provide evidence for additional func-
tions of Grx2 in the cytosol, in cell proliferation, and in cellular differentiation. Antioxid. Redox Signal. 11, 1–14.
1
Introduction
G
stasis. It was first discovered as a hydrogen donor for ri-
bonucleotide reductase (RNR) in Escherichia coli (12) but has
since been identified in a wide range of organisms, includ-
ing eukaryotes, prokaryotes, and viruses. Grx effectively cat-
alyzes the reduction of both glutathione-mixed disulfides
and protein disulfides by using one or two redox-active cys-
teines in its active site, respectively [for an overview, see (4)
and (19)]. This active-site motif, Cys-X-X-Cys, is character-
istic among a growing number of redox active enzymes that
constitute the thioredoxin family of proteins. Thioredoxins
(Trxs) and Grxs share a similar three-dimensional structure
LUTAREDOXIN (Grx) is a small ubiquitous protein that is
involved in the maintenance of cellular redox homeo-
called the thioredoxin fold (28) and reduce disulfides with
electrons provided from NADPH via thioredoxin reductase
(TrxR) and glutathione/glutathione reductase, respectively
(13). Complete Trx and Grx systems are present in both cy-
tosol and mitochondria (8, 24, 30). Three different mam-
malian Trx systems have been described: the cytosolic
Trx1/TrxR1, the mitochondrial Trx2/TrxR2, and thioredoxin
and glutathione reductase (TGR). TGR differs from the two
other TrxRs because of an N-terminal Grx domain. It is ex-
pressed predominantly in testis, has a broad substrate speci-
ficity, and can reduce components from both the Trx and Grx
systems (45, 46).
At present, two different Grxs have been characterized in
mammals. Cytosolic Grx1 contains the active-site sequence
Cys-Pro-Tyr-Cys and is involved in various redox-depen-
1Medical Nobel Institute for Biochemistry, Department of Medical Biochemistry and Biophysics, Karolinska Institutet, Stockholm, Swe-
den.
2The Institute for Clinical Cytobiology and Cytopathology, Phillips Universität, Marburg, Germany.
3The Department of Biochemistry, School of Medicine, University of Buenos Aires, Buenos Aires, Argentina.

Page 2

dent cellular processes, such as reduction of ribonucleotides
and dehydroascorbate (12, 50), actin polymerization (49), and
protection against oxidative stress and apoptosis (5, 32); for
a detailed overview, see (4, 19, 43)]. The second mammalian
Grx (Grx2), localized predominantly in mitochondria, was
identified more recently (8, 24). The protein levels of total
Grx2 are significantly lower compared with the levels of
Grx1 in placenta and cell lysates from various tumor cell lines
(23). It was suggested that Grx2 plays a central role in the
interplay between GSH and the thiol groups of mitochon-
drial membrane proteins (2). Transient silencing of Grx2 ex-
pression in HeLa dramatically sensitized these adenocarci-
noma cells to apoptosis induced by phenylarsine oxide and
the anticancer agent doxorubicin (21). Overexpression of
Grx2 attenuates apoptosis induced by doxorubicin and the
antimetabolite 2-deoxy-D-glucose by sustaining mitochon-
drial redox homeostasis, thereby preventing cardiolipin ox-
idation, cytochrome c release, caspase activation, and ulti-
mately, cell death (7).
The active-site sequence of Grx2 (Cys-Ser-Tyr-Cys) differs
from the Grx consensus sequence, with the Pro being ex-
changed for Ser. This exchange has two remarkable conse-
quences. First, Grx2 can be reduced not only by GSH, but
also by thioredoxin reductase (16). The physiologic impor-
tance of this reaction remains to be established; however, this
electron pathway might play a role under conditions in
which the GSH/GSSG ratio decreases or the cellular pH de-
creases or both. Second, it allows the protein to complex a
2Fe2S cluster (3, 20). This cluster bridges two Grx2
monomers to form a dimeric holo-Grx2 complex. The chro-
mophore is complexed by the two N-terminal active-site thi-
ols of the proteins and the thiol groups of two molecules of
GSH non-covalently bound to the Grx2 monomers (3, 15).
The holo-complex lacks enzymatic activity; dissociation of
the cluster yields active monomeric apoprotein. Oxygen-de-
pendent degradation of the holo-complex is efficiently pre-
vented by reduced GSH. GSSG and other redox-active com-
pounds conversely promote degradation (20).
The human Grx2 gene (GLRX2) is located at chromosome
1q31.2-31.3 (24) and gives rise to three alternatively tran-
scribed Grx2 mRNA isoforms with alternative first exons (22,
24). These mRNA variants encode three distinct proteins: (a)
the ubiquitous Grx2a (exons Ia–IV) that is targeted to mito-
chondria; (b) Grx2b (exons Ib–IV); and (c) Grx2c (exons
Ibtruncated–IV) share a cytosolic and nuclear localization pat-
tern. Their expression is restricted to testis and cancer cells
(22). In mouse, mRNA variants of Grx2 differing in the 5?
end have been proposed (8). One of these variants contains
a mitochondrial signal peptide and corresponds to the hu-
man mitochondrial isoform, Grx2a. However, no counter-
part for the proposed alternative exon Ib, which gives rise
to the cytosolic/nuclear isoforms Grx2b and Grx2c, is pres-
ent in the mouse gene. Analyzing the absolute gene-expres-
sion pattern of Grx2 in mouse, Jurado et al.(17) found that
Grx2a mRNA accounted for only 40% of the total amount of
Grx2 mRNA in most organs. The remarkable exception to
this were testes, in which only 1% of the total Grx2 mRNA
accounted for the Grx2a variant. These studies strongly in-
dicated the presence of further Grx2 transcript variants in
mouse as well. Here, we present the identification of these
additional transcript variants, their expression pattern in
multiple tissues, and a characterization of a new nonmito-
chondrial Grx2 protein isoform that is present in many tis-
sues.
Material and Methods
Materials and general methods
Chemicals and enzymes were purchased from Sigma, un-
less otherwise stated, and were of analytic grade or better.
C57BL/6J mice were purchased from the Jackson Laboratory
(Bar Harbor, ME). All experiments involving animals were
approved by the respective authorities. SDS-PAGE was run
by using the Novex Mini-Cell and precast NuPAGE gels
(12% acrylamide, Bis-Tris, Invitrogen, Carlsbad, CA) ac-
cording to the manufacturer’s instructions.
Analysis of EST sequences
Based on the sequences of the proposed mouse glutaredoxin
2 exons Ia, II, III, and IV (e.g., GeneBank accession number
AF380337), an extensive EST cluster search was carried out.
Each exon was compared with the GenBank mouse EST data-
base by using the BlastN algorithms (1). Sequences with down
to 40-bp matches were analyzed in detail. When necessary, se-
quences were reoriented, and each EST sequence was individ-
ually mapped to the genomic sequence (AL592403) by using
EST2genome from the EMBOSS package (http://emboss.
sourceforge.net) to identify potential exon–intron boundaries.
cDNA and PCR
For the identification of transcript variants and analysis of
expression patterns, multiple tissue cDNA panels I and II
(Clontech, Mountain View, CA) were purchased, which in-
cluded primers for the amplification of glyceraldehyde-3-
phosphate dehydrogenase (G3PDH), which served as posi-
tive control. Additional cDNA was produced from C57BL/6J
mice. RNA isolation was done by using RNeasy Mini Kit (Qi-
agen, Hilden, Germany) followed by immediate first-strand
cDNA synthesis by using the Omniscript RT-Kit (MBI Fer-
mentas, St. Leon-Rot, Germany) as suggested by the manu-
facturer by using an oligo-dT primer and 3 ?g of RNA.
Oligonucleotides for PCR (summarized in Table 1) were or-
dered from DNA technology (Aarhus, Denmark). Reactions
were performed in programmable thermocyclers from MJ
Research in a total volume of 20 ?l containing 1.25 U re-
combinant Taq polymerase, 200 ?M dNTPs, 1.5 mM Mg2?
(all MBI Fermentas), 0.5 mM of each primer, and ?4 ng first-
strand cDNA. After an initial denaturation step at 95°C for
2 min, 35 cycles of the following program were performed:
95°C for 30 sec, reaction-specific annealing temperatures (see
Table 1) for 45 sec, and 72°C for 40 sec. The final elongation
step was performed at 72°C for another 10 min. For tran-
script variant mGLRX_v2, we used EST-clone P998004867
from the I.M.A.G.E. Consortium as positive control. In gen-
eral, PCR products were analyzed on 1.5% agarose gels
(14.5 ? 19 cm containing 40 mM Tris, 19 mM acetic acid, and
1 mM EDTA). Selected PCR fragments were extracted from
agarose gels by using the Qiagen Gel Extraction Kit and di-
rectly ligated into pGEM-T vector (Promega, Madison, WI).
The plasmids were subsequently sequenced from both 3?and
5? directions by the core facility of the Karolinska Institute,
KISeq. The densiometric analysis of band intensities was per-
formed by using the Kodak 1D software, version 3.5.
HUDEMANN ET AL.2

Page 3

Rapid amplification of cDNA ends (5?RACE)
To obtain the 5? end of the newly identified Grx2 mRNA
variants, the GeneRacer kit (Invitrogen) was used, according
to the manufacturer’s recommendations. In brief, 2.5 ?g to-
tal RNA from mouse testis were dephosphorylated and de-
capped. The RNA Oligo was ligated to the 5? end of the
mRNA and subsequently reverse transcribed with the Gen-
eRacer Oligo dT Primer by using SuperScript III Reverse
Transcriptase (Invitrogen). 5?-RACE reactions were per-
formed by using the gene specific primer 5?-cac tga tga acc
aga ggc agc aat ttc-3? and the GeneRacer 5? Primer. For each
reaction, 2 ?l of cDNA and 0.5 ?l of DyNAzyme EXT DNA
polymerase (Finnzymes) were used to amplify the fragment
by touchdown PCR method by using the following condi-
tions: One cycle at 94°C, 2 min; 10 cycles at 94°C, 15 sec; 76°C,
30 sec, minus 0.5°C each cycle, and 72°C, 1 min; Thirty cy-
cles at 94°C, 15 sec; 71°C, 30 sec; and 72°C, 1 min plus 5 sec
for each cycle; and final extension at 72°C for 7 min. After
amplification, samples were kept at 4°C until analysis. The
5?-RACE fragment was gel-purified and cloned into pCR4-
TOPO vector (Invitrogen). Positive clones were analyzed by
PCR and sequencing.
Cloning, recombinant expression, and purification of
mouse Grx2 isoforms
The open reading frames of mouse Grx2c and Grx2d were
amplified from pooled cDNA by using the following oligonu-
cleotides: 5?
for both reactions:
GGGAAACAGCACATCGTCG, 3? reverse for Grx2c: CAC-
ACAGGATCCTCAATGTCTTTCCTCTTGTTTTTTTTTTAAA
TAACAC, and 3?
reverse
GGATCCTCAGGCAACTACTGAGATTTTTTTCAG.
CACACACATAT-
for Grx2d: CACACA-
The
PCR products were ligated into the NdeI and BamH1 sites of
pET15b (Novagen, Darmstadt, Germany). Human Grx2c and
mouse Grx2c were essentially expressed and purified as de-
scribed earlier (20), replacing nickel nitrilotriacetic acid columns
with Talon columns (Clontech) of 5 ml. Bound protein was
eluted with a 200-ml linear gradient of imidazole from 0 to 300
mM in 50 mM sodium phosphate, pH 8, containing 300 mM
NaCl. Preparation of apo Grx2c was done with high-perfor-
mance liquid chromatography, as described in reference (20).
Grx2d was not expressed in soluble form. Therefore, we
used a modified protocol for in vitro folding of inclusion
body proteins (38). As described earlier, cells were disrupted
by a combination of lysozyme treatment and sonication. The
crude extract was centrifuged at 20,000 g for 30 min. The pel-
let containing Grx2d inclusion bodies was resuspended in 50
mM sodium phosphate buffer, pH 8, 300 mM NaCl, and 1%
Triton-X100 and centrifuged at 12,000 g for 30 min. The pel-
let was resuspended in 50 mM sodium phosphate buffer, pH
8, containing 300 mM NaCl, and centrifuged in the same
way. Next, the pellet was resuspended in 6 Mguanidine HCl,
100 mM dithiotreitol, and 100 mM Tris/HCl, pH 8, and in-
cubated for 2 h at room temperature with gentle shaking.
The protein solution was centrifuged at 20,000 g for 30 min.
The supernatant was collected, and dithiothreitol was re-
moved by using prepacked Sephadex G-25 gel filtration
columns (PD10, Amersham, Uppsala, Sweden) equilibrated
with 6 M guanidine HCl and 100 mM Tris/HCl, pH 8. The
protein was purified by immobilized metal-affinity chro-
matography by using a Talon column of 3 ml. Unspecific
bound protein was removed by two wash steps with 50 mM
sodium phosphate, pH 8, 300 mM NaCl, 6 M guanidine HCl,
and the same buffer containing 50 mM imidazole. The pro-
tein was eluted in the washing buffer containing 300 mM
ISOFORMS OF MOUSE Grx23
TABLE 1.OLIGONUCLEOTIDES AND POLYMERASE CHAIN REACTIONS
Oligonucleotides
NameTarget exonSizeSequence
PR89
PR108
PR119
PR90
PR91
PRCH6
PR105
PR93
PR94
Ia
Ia/Iatrunk
Ic/Ictrunk
Id
II
IIIb (reverse)
IIIb
IV (reverse)
IV-UTR (reverse)
25
18
29
28
30
27
27
27
28
5?-GCTGGTGGCGAGCGGGAGGATCTTG-3?
5?-CTGCTCTCCCGGAGGCTG-3?
5?-CTTAAGATGACATTCCGGTCCGGTCCATC-3?
5?-GTGCCTGCCCCTTTCCTTAATAGGAAAG-3?
5?-CAGCACATCGTCGTTTTGGGGGAAGTCTAC-3?
5?-TCAGGCAACTACTGAGATTTTTTTCAG-3?
5?-CTCAGTAGTTGCCTGAGGGAGAGAATG-3?
5?-CACTGATGAACCAGAGGCAGCAATTTC-3?
5?-CAGCCTCAAGTCAAAGGTACGACTGCAC-3?
Polymerase chain reactions
Oligonucleotides Target sequenceExpected length Annealing temperature Positive control
G3PDH Primermix
PR91/PR94
PR89/PR94
PR119/PR94
PR90/PR94
PR108/PR93
PR108/PRCH6
PR105/PR94
G3PDH for each tissue
Exon II-IV UTR rev
Exon Ia-IV UTR rev
Exon Ic/Ictrunc-IV UTR rev
Exon Id-IV UTR rev
Exon Ia/Iatrunc-IV rev
Exon Ia/Iatrunc-IIIb rev
Exon IIIb-IV UTR rev
983
430
502
55
58
58
56
54
54
54
55
Pooled cDNA
Pooled cDNA
Pooled cDNA
Pooled cDNA
None available
EST clone P998004867
None available
None available
658/526
502
471/397
365
269

Page 4

imidazole. The pure protein was stepwise diluted 12-fold with
50 mMsodium phosphate, pH 8, 500 mMNaCl containing 50%
glycerol over a period of 4 h on ice to allow gentle refolding of
the protein. Residual guanidine and imidazole were removed
by using PD10 columns equilibrated with the dilution buffer.
Protein concentrations were determined spectrophoto-
metrically by using the absorbance coefficients of 11,460
M/cm for Grx2c and 9,870 M/cm for Grx2d at 280 nm. These
absorbance coefficients were calculated from the protein se-
quences by using ProtParam (www.expasy.ch).
ELISA
Organ tissue from two different mice (except for liver, for
which only one organ was used) were homogenized in 400
?l lysis buffer (10 mM Tris/HCl, pH 7.4, 10 mM NaCl, 3 mM
MgCl2, 0.1 % Nonidet P40) by using pellet pestles (Kontes
Glass Company, Vineland, NJ) in an Eppendorf tube. After
15 min at room temperature, the samples were frozen in liq-
uid nitrogen. After thawing, the samples were centrifuged
at 20,000 g for 60 min at 4°C. The total protein concentration
in the supernatant was determined by using Bradford
Reagent (BioRad, Hercules, CA) with a BSA-standard from
0.01 to 0.8 mg/ml. Quantification of mouse Grx2 was done
by using a specific sandwich ELISA, as described by Lund-
berget al. (23). The antibody used in this ELISA was raised
against recombinant human Grx2c and thus recognizes all
isoforms. Human Grx2c was exchanged for mouse Grx2c to
generate the standard curve used for quantification.
Immunohistochemistry and immunofluorescence
Mouse testes were dissected and fixed for 24 h in either
Carnoy (immunofluorescence) or 3.7% formalin (immuno-
histochemistry). After dehydration, the testes were
processed for paraffin embedding. Sections (10 ?m thick)
were cut and placed on poly-L-lysine–coated slides. The sec-
tions were deparaffinized and incubated in 3% hydrogen
peroxide for 30 min to quench endogenous peroxidase. Af-
ter three washes in PBS, nonspecific antibody-binding sites
were blocked with 5% normal goat serum (Invitrogen). Next,
sections were incubated overnight with the primary anti-
body, rabbit anti-mouse Grx2 (diluted 1:1,000), at 4°C. The
mouse Grx2 antibody was generated as described by Lund-
berg et al. (24) by using recombinant mGrx2c and detects
Grx2a, processed Grx2a, and Grx2c. Negative controls were
incubated with goat serum in PBS. The following day, sec-
tions were washed with PBS and subsequently incubated
with a biotinylated goat anti-rabbit antibody diluted 1:500
(Invitrogen, Karlsruhe, Germany) for 30 min at room tem-
perature. For immunohistochemistry, the StreptABCom-
plex/HRP (Dako Cytomation, Hamburg, Germany) was
used for antigen staining, according to the manufacturer’s
recommendations. The sections were incubated with the sub-
strate aminoethyl carbazole (AEC; Invitrogen, Karlsruhe,
Germany) for 5 min at room temperature, counterstained
with Mayer’s hematoxylin, and mounted with Mowiol. For
immunofluorescence studies, the TSA kit (Tyramide Signal
Amplification; Invitrogen) was used. In brief, after incuba-
tion with the biotinylated secondary antibody, the sections
were incubated with a streptavidin-conjugated horseradish
peroxidase for 30 min at room temperature. After three PBS
washing steps, sections were incubated with the Alexa Fluor
488–labeled tyramide for 5 min at room temperature. Sec-
tions were counterstained with Hoechst 33342 (Sigma).
The cellular localization of Grx2 in mouse tissues was an-
alyzed by fluorescent co-staining with cytoplasmatic (DNase
I, Invitrogen) and mitochondrial markers (anti-prohibitin 1,
Calbiochem or anti-manganese superoxide dismutase (Mn-
SOD, Alexis). Instead of the signal-amplification kit de-
scribed earlier, an antigen-recovery step was introduced by
heating the slides in citrate buffer (10 mM, pH 6.0) for 14
min in a microwave oven. After incubation overnight with
the anti-mouse Grx2 (1:200) and the anti-prohibitin 1 (1:200)
or anti-MnSOD (1:200) antibodies, the slides were developed
with goat anti-rabbit Alexa Fluor 633 (1:300), detecting anti-
Grx2, and chicken anti-mouse Alexa 488 (1:500, both Invit-
rogen), detecting anti-prohibitin 1 or anti-MnSOD. Alexa
Fluor 488–labeled DNase I was applied at 9 ?g/ml for 20
min for the labeling of globular actin. DNA was counter-
stained with Hoechst as described earlier; the slides were
mounted with Mowiol.
Sections prepared for immunohistochemistry were exam-
ined with a Leica Diaplan microscope equipped with a Mi-
croPublisher camera (QImaging, Surrey, BC, Canada). Sam-
ples prepared for immunofluorescence were analyzed with
a Leica TCS SP2 confocal laser scanning microscope by us-
ing a 40? oil planapochromatic lens (Leica, Heidelberg, Ger-
many). Deconvolution and 3D reconstruction was performed
by using the software package Huygens (Scientific Volume
Imaging, Hilversum, The Netherlands).
Analysis of chromophore and quaternary structure
UV-visible spectra were recorded with a Shimadzu UV-
2100 spectrophotometer. The presence of dimeric Grx2c and
the molecular weights were confirmed with high-perfor-
mance liquid chromatography (Amersham Smart system) by
using a Superdex 75 3.2/30 precision column equilibrated
with 50 mM sodium phosphate, pH 8, and 300 mM NaCl, as
described in (20). The Low Molecular Weight Gel Filtration
Calibration Kit (Amersham) was used for calibration.
Enzymatic assay
Grx2 activity was measured by using a modified protocol
of the hydroxyethyl disulfide (HED)-assay (25) in a total vol-
ume of 200 ?l. In brief, 50 ?l HED (0.7 mM final concentra-
tion) was added to 100 Êl of a freshly prepared substrate mix
containing 100 mM Tris/HCl, pH 8.0, 2 mM EDTA, 0.1
mg/ml BSA, 1 mM GSH, 240 ?M NADPH, and 6 ?g/ml
yeast glutathione-reductase (final concentrations). After 4
min of preincubation at 30°C, different amounts of Grx in a
volume of 50 ?l (0,10, 25, 50, 75, 100, 150, and 200 nM final
concentration, respectively) were added. The decrease in ab-
sorbance at 340 nm was measured at 30°C in a 96-well plate
by using a VERSAmax microplate reader (Molecular De-
vices). NADPH consumption was quantified based on 12
standards of NADPH from 0 to 240 ?M concentration. Ac-
tivity was expressed as micromoles of NADPH oxidized per
min.
RNR activity was determined by following the conversion
of [3H]CDP into [3H]dCDP with mouse RNR, according to
the methods of Thelander and co-workers (6, 27). Reducing
equivalents for Grx2c were provided through 4 mM GSH,
glutathione reductase, and NADPH (25). The reaction was
HUDEMANN ET AL.4

Page 5

initiated by adding mouse RNR (5.5 ?g of R1 and 2.1 ?g of
R2) to the reaction mixture containing 40 mM Tris-Cl buffer,
pH 7.5, 2 mM ATP, 10 mM magnesium chloride, 200 mM
potassium chloride, and 0.5 mM [3H]CDP (23,000 cpm/
nmol) in a final volume of 50 ?l. Samples were incubated at
37°C, and the reaction was terminated after 30 min by the
addition of 1 M HClO4followed by acid hydrolysis. The
amount of [3H]dCDP formed was calculated from the results
of liquid scintillation counting after ion-exchange chro-
matography separation of the monophosphates on Dowex-
50 columns.
Nomenclature
We have used the official Grx2 gene name GLRX2 when
discussing properties of the gene. Transcript variants were
named as proposed by the International Committee on Stan-
dardized Genetic Nomenclature for Mice (http:/ /www.in-
formatics.jax.org/mgihome/nomen/gene.shtml). The com-
monly used abbreviation Grx was used to describe the
proteins. To avoid any possibility of confusion, we have kept
and extended the nomenclature for Grx2 exons and isoforms
introduced by Lundberget al. (24). The newly identified al-
ternative first exon was named exon Ic to avoid confusion
with human exon Ib, not present in mouse. The new protein
variants were named Grx2c and Grx2d.
Results
Identification and expression pattern of
Grx2 transcript variants
The gene encoding mouse Grx2 (GLRX2) is located on
chromosome 1 (1 77.8 cM) and consists of four exons span-
ning ?12 kb. We systematically screened the mouse genome
database for EST sequences homologous to at least one of
the Grx2 proposed exons (8) and identified ?100 unique en-
tries. We aligned these potential mRNA variants pairwise to
the genomic sequence, analyzed the genomic structure, and
identified a variety of potential exons and splice variants. To
verify the additional exons and variants and to analyze the
expression pattern of Grx2 transcript variants, we screened
mouse cDNA derived from heart, brain, spleen, lung, liver,
skeleton muscle, kidney, testis, embryo day 7, 11, 15, and 17,
placenta, eye, lymph node, smooth muscle, prostate, thymus,
stomach, and uterus with PCR (Fig. 1). Selected PCR prod-
ucts were cloned and sequenced to confirm the exons and to
reveal intron/exon junctions.
Transcript variants encoding functional Grx2 isoforms
(i.e., transcripts that encode all residues necessary for enzy-
matic activity and the Trx fold) consist of at least exons II,
III, and IV. These transcripts were found to be expressed in
all tissues. In addition, we identified a second ?120-bp larger
product in testis (Fig. 1, second panel). Mitochondrial Grx2a
(transcript variant 1, mGLRX2_v1), consisting of exons Ia, II,
III, and IV, was detected in all tissues with the exception of
testis, which is in agreement with earlier results (17) (Fig. 1,
third panel). For the identification of an alternative splice-
donor site within exon Ia implied by our EST analysis, we
established a PCR that amplified both potential transcripts.
We identified a truncated exon Iatruncthat is 73 bp shorter
than exon Ia. Because of this difference in size, we were able
to distinguish between the two forms (Fig. 1, panel four).
Subsequent sequencing confirmed the presence of this tran-
script variant (named mGLRX2_v2) in many tissues. Esti-
mated from the density of the bands, an average of 80% of
the mGLRX2_v1/mGLRX2_v2 transcripts contained exon Ia,
ISOFORMS OF MOUSE Grx25
FIG. 1.
tissues and pooled cDNA. Ten microliters of each PCR reaction was run on a 1.5% agarose gel (except Iatrunc-IV ? 2%
agarose) containing 0.01% ethidium bromide in Tris-acetate-EDTA buffer. Negative control contained water as template,
positive pooled cDNA. EST clone P998004867, used as positive control for transcript variant Iatrunc-IV, was derived from
the I.M.A.G.E. Consortium. A G3PDH oligonucleotide pair was used for cDNA integrity check in each tissue. Oligonu-
cleotides and PCR conditions are summarized in Table I.
RT-PCR analysis of the expression of different mGrx2 transcript variants in cDNA derived from different mouse

Page 6

FIG. 2.

Page 7

and 20%, Iatrunc. Again, we obtained an additional larger
product from testis cDNA. Sequencing of the larger prod-
ucts from the variant 2- and 3-specific PCR reactions as well
as from the PCR specific for the common part of all variants
(exons II–IV) revealed the third alternative transcript variant
(mGLRX_v3). This variant consists of exons Iatrunc, II, III, the
newly identified single-cassette exon IIIb, and exon IV. We
screened for variant mGLRX2_v3 amplifying fragments from
both exons Iatruncto IIIb and exons IIIb to IV. The PCRs re-
sulted in strong bands in testis samples, but only faint and
inconsistent bands in independent samples derived from
other tissues (Fig. 1, panels five and six).
As implied by the EST screening, we identified a new al-
ternative first exon in the mouse GLRX2 gene. This exon is
located upstream of exon Ia in a position similar to that of
the human exon Ib; however, the sequences do not share sig-
nificant homology. Our PCR screening for exon Ic resulted
in the identification of transcript variants four and five. Be-
cause of two alternative splice donors in exon Ic,
mGLRX2_v4 consists of exons Ic, II, III, and IV, mGLRX2_v5
contained a 132-bp shorter form of exon Ic, Ictrunc, and ex-
ons II, III, and IV. Both forms are expressed predominantly
in testis. The ratio between variants four and five in testis
was found to be 2:1. By using 5?-RACE, we determined the
transcription start point of mGLRX2_v4 and mGLRX2_v5 to
be located 354 bp upstream of the splice-donor site of
mGLRX2_v4 (Fig. 2B).
All experimentally verified mGLRX2 exons and transcript
variants are summarized in Table 2 and Fig. 2. The five iden-
tified transcript variants encode three distinct proteins: (a)
mGLRX2_v1 encodes the mitochondrial isoform Grx2a; (b)
mGLRX2_v2, mGLRX2_v4, and mGLRX2_v5 share the same
open reading frame and give rise to a protein entirely en-
coded by the constitutive exons II, III, and IV that is essen-
tially identical to human cytosolic/nuclear Grx2c; and (c)
mGLRX_v3 encodes the second new Grx2 isoform (named
ISOFORMS OF MOUSE Grx27
TABLE 2.OVERVIEW OF GLRX2 EXONS AND TRANSCRIPT VARIANTS
Exons
Size (coding region)
Bp
Intron size
bpExon EncodesSplice acceptorSplice donor Intron
Ia
Iatrunc
Ic
Ictrunc
II
III
98ATG, mitochondrial translocation signal
(-)
(-)
(-)
ATG, 1st nonactive site Cys
Active site, GSH binding
(-)
(-)
(-)
(-)
ttagAATG
ttagGAAA
CGGGgtac
TTCGgtgg
GCAGgtga
TCAGgtaa
CCAAgtga
AACCgtga
Ia-II
Iatrunc-II
Ic-II
Ictrunc-II
II-III
III-IIIb
III-IV
IIIb-IV
(-)
1889
1963
2205
2337
3335
758
1284
413
(-)
0
0
0
64
177
IIIb
IV
45
132
Alternative TGA
cis-Pro, GSH binding, 2nd nonactive site
Cys, TGA
caagGACA
acagGTTC
CAAGgtca
(-)
Transcript variants
Open reading
frame
Kozak
sequencea
Encodes (protein
isoform)
Amino
acidsName Exons in mRNA Expressed in (tissues)
GLRX2_v1 Ia, II, III, IV Ia-IVTCGGGAATGG Grx2a (mitochondria) 156All tissues, very low in
testis
All tissues
Testis
Testis
Testis
GLRX2_v2 Iatrunc, II, III, IV
GLRX2_v3 Iatrunc, II, III, IIIb, IV
GLRX2_v4 Ic, II, III, IV
GLRX2_v5 Ictrunc, II, III, IV
II-IV
II-IIIb
II-IV
II-IV
CTTCGAATGG
CTTCGAATGG
GGCAGAATGG Grx2c (cytosol)
CTCAGAATGG
Grx2c (cytosol)
Grx2d (cytosol)
123
96
123
123 Grx2c (cytosol)
aBold characters indicate consistencies with the general Kozak consensus sequence.
FIG. 2.
and transcript variants of mouse GLRX2 (lower panel). The longest possible open reading frames are indicated by arrows
(start codons, translation initiation) and hooks (stop codons). (B) Sequence of exons Ic and Ia. The transcription start points
are marked with triangles; the exons sequence are shown in bold-italic characters, experimentally determined splice donor
sites are boxed, and the coding region of exon Ia is underlined. (C) Comparison of the primary structures of mouse Grx2a,
Grx2c, and Grx2d. Some structurally and functionally important regions are indicated above the sequences. Residues im-
portant for GSH binding were printed on black, and the cysteines forming the structural disulfide on gray background. (D)
Comparison of the human and mouse GLRX2 gene structures and alternative splicing/transcription initiation patterns.
Note: With the exception of the 5? of mouse exon Ic, homologous to a region immediately upstream of human exon Ib,
mouse exon Ic and human exon Ib do not share significant sequence homology. The regions of highest homology are (in
order of their degree of identity) exons III, exons Ia, exons IV, and exons II. Additional regions of high homology are cen-
tered in the middle of both introns II and III.
Summary of mouse GLRX2 transcript variants. (A) Alternative splicing of the mouse GLRX2 gene (upper panel)

Page 8

Grx2d) that is encoded by an open reading frame spanning
from exon II to exon IIIb.
Properties of the mouse Grx2 protein isoforms
The newly identified cytosolic isoforms Grx2c and Grx2d
were expressed in E. coli and purified for further character-
ization. Similar to the human counterpart (20), purified
mouse Grx2c was not clear but appeared brownish. In gel-
filtration chromatography, the protein (theoretic mass with
6-His tag, 16.5 kDa) separated into two fractions: a brown-
ish dimeric fraction (apparent size, 36.4 kDa) and a clear
monomeric fraction of 14.2 kDa (Fig. 3). The dimeric fraction
exhibited the same spectral properties as human Grx2 (i.e.,
two additional broad absorption bands at 320 and 428 nm)
(Fig. 3). Similar to human Grx2, the monomeric fraction of
mouse Grx2c was enzymatically active, with a specific ac-
tivity of 11.3 ? 1.3 ?mol/mg/min in the HED assay com-
pared with 27.4 ? 2.3 ?mol/mg/min for human Grx2c. To
study a possible function of Grx2c in the cytosol, we tested
the apoprotein for activity with mouse RNR. As shown in
Fig. 4, Grx2c was able to donate electrons to RNR for the re-
duction of CDP, yielding a Vmaxof 0.43 nM dCDP/min ?
0.04. The kmof RNR for Grx2 was determined to be 0.15
?M ? 0.05.
Unlike Grx2c, Grx2d was difficult to obtain, because it was
not expressed as soluble protein in E. coli. This was not un-
expected because the protein lacks some residues crucial for
the thioredoxin fold, most notably the cis-proline residue
(Fig. 2C). We purified Grx2d from inclusion bodies in the
presence of guanidine hydrochloride and allowed the pro-
tein to refold in the presence of 500 mM NaCl and 50% glyc-
erol at pH 8. According to CD spectroscopy, the in vitro
folded protein exhibited a similar content of helices, sheets,
and coiled regions as Grx2 (data not shown). However, the
protein was inactive in the HED assay, even when assayed
at high concentrations (i.e., up to 0.5 ?M).
Immunologic detection of Grx2 in mouse tissues
First, we compared the reaction of mouse Grx2c and Grx2d
with the antibodies obtained against human Grx2 in a spe-
cific sandwich ELISA (23). The standard curve for mouse
Grx2c was colinear to the standard for human Grx2; how-
ever, the sensitivity was about sixfold lower. In vitro folded
Grx2d did not react with the antibodies in a specific manner
(Fig. 5A). Of course, this does not completely exclude that
natively folded Grx2d may be recognized by the antibody.
Next, we used ELISA to confirm and quantify the presence
of Grx2 (i.e., most likely the sum of Grx2a and Grx2c, in to-
tal extracts of different tissues) (Fig. 5B). Heart contained the
highest amount of immunoreactive Grx2, followed by kid-
ney, brain, testis, spleen, and liver. Because GLRX2_v1 (mi-
tochondrial Grx2a) is basically absent from testis, these re-
sults provided further evidence for the presence of
nonmitochondrial Grx2 in testis tissue.
Next, we studied the localization of Grx2 in mouse testis
cell types with immunohistochemistry and confocal im-
munofluorescence microscopy in greater detail (Fig. 6). Grx2
immunolabeling resulted in a consistent and strong staining
in spermatogonia that was concentrated in the cytoplasm
(Fig. 6A and B). Interstitial Leydig cells showed some degree
of staining, whereas all stages of spermatocytes and Sertoli
cells showed only faint staining (Fig. 6A). Spermatids dis-
played the strongest Grx2 staining. Round spermatids
showed a distinct cytosolic staining concentrated in one
hemisphere of the cell (Fig. 6C and D). The 3D reconstruc-
tion of 62 layers within a 10-?m section revealed a concave,
cuplike distribution of the protein (Fig. 6E). Staining of Grx2
in elongated spermatids resulted in an arrowhead-like struc-
ture on top of the elongated nuclei, pointing away from the
lumen of the tubulus (Fig. 6F–H). The staining in both round
and elongated spermatids is in agreement with acrosomal
localization.
HUDEMANN ET AL.8
FIG. 4.
bonucleotide reductase. Michaelis–Menten plot of reaction
velocity versus substrate concentration.
Mouse Grx2c as electron donor for mouse ri-
FIG. 3.
sorption spectra of human Grx2 (dashed line), mouse Grx2c
(straight line), and mouse Grx2d (dotted line). The spectral
bands at 320 and 428 nm are in agreement with a 2Fe2S clus-
ter, as demonstrated for human Grx2 before (20). Inset: Elu-
tion profile of mouse Grx2c purified on a Superdex 75 gel-
filtration column at 280 nm (protein) and 320 nm ([Fe,S]
chromophore). The determined molecular weights were 36.4
kDa for the colored dimeric protein and 14.2 kDa for the
monomeric protein.
FeS cluster binding to mouse Grx2. UV-visible ab-

Page 9

FIG. 6. Localization of Grx2 in mouse testis. (A) Light-microscopic localization of Grx2 revealed by staining with anti-mouse Grx2
antibodies. (B) Magnification of A (yellow frame) showing intense staining in the cytosol of spermatogonia (SG). (C–E) Staining of
round spermatids (RSs). (C) Immunohistochemical staining of RS. (D) Confocal immunofluorescence detection of Grx2 in RS. (E) 3D-
reconstruction of 62 layers of D spanning 10 ?m. (F–H) Staining of elongated spermatids (ESs). (F) Immunohistochemical staining of
ESs. (G) Confocal immunofluorescence detection of Grx2 in ESs. (H) Co-staining of Grx2 (red) with the cytosolic marker DNase I
(green) and the nucleus (blue), analyzed with confocal microscopy. Further abbreviations: spermatocyte (SC) and Leydig cell (LY).
FIG. 5.
(A) ELISA standard curves of hu-
man Grx2 (circles), mouse Grx2c
(squares), and mouse Grx2d (dia-
monds) by using the sandwich
ELISA described in (23). (B) Levels
of Grx2 in mouse tissue by using
mouse Grx2c as standard for quan-
tification.
Levels of Grx2 in tissues.

Page 10

Further to investigate the presence of cytosolic Grx2 in
cells, we developed a method for the co-staining of Grx2 with
cytosolic (Dnase I for the staining of globular actin) and mi-
tochondrial (anti-prohibitin 1 and anti-manganese superox-
ide dismutase) markers and analyzed the subcellular local-
ization of Grx2 in different cells of the testis, the spleen, and
the stomach, tissues that all exhibit a high amount of tran-
script encoding Grx2c (see Fig. 1). As expected, spermato-
gonia show a clear co-staining of Grx2 with the cytosolic (Fig.
7A) but not the mitochondrial marker (Fig. 7B). In stomach,
parietal cells displayed mitochondrial co-staining with Grx2
(not shown). In enteroendocrine cells of the fundic glands,
however, Grx2 does not co-stain with either of the mito-
chondrial marker proteins investigated (Fig. 7C and D). In
spleen, cells of the red pulpa show a clear co-staining of Grx2
with mitochondria (Fig. 7E and F). In cells of the white pulpa,
conversely, cells display mainly cytosolic localization of Grx2
(Fig. 7G and H). These results demonstrate the presence of
cytosolic Grx2, presumably Grx2c, in specialized cells of dif-
ferent tissues, implying specific differences in Grx2 tran-
scription and splicing in different cell types within a given
tissue, at least in the three tissues investigated.
Discussion
Both the human and mouse GLRX2 genes are, by mecha-
nisms of alternative transcription initiation and splicing,
transcribed and processed into different mRNA variants that
encode functionally different proteins. Alternative use of the
genome is one of the striking differences between related
species, and a number of similarities but also remarkable dif-
ferences are noted between human and mouse Grx2 mRNA
variants and protein isoforms (Fig. 2A and D).
The human GLRX2 gene consists of five exons: the con-
stitutive exons II, III, and IV and the two alternative first ex-
ons Ia and Ib (22). Transcript variant 1 (hGLRX2_v1) is com-
posed of exons Ia–IV and encodes Grx2a, hGLRX_v2 of exons
Ib – IV (Grx2b), and alternative splicing of exon Ib yields
hGLRX2_v3 (Grx2c). The ubiquitous Grx2a is a mitochondr-
ial protein, and the testis-specific Grx2b and c are localized
in both cytosol and nucleus. In the present study, we iden-
tified, confirmed, and characterized six exons in the mouse
Grx2 gene (Fig. 2A). The three constitutive exons II, III, and
IV, two alternatively transcribed first exons (exons Ia and Ic),
which both contain two alternative splice-donor sites, and
one single-cassette exon (exon IIIb), which resides between
exons III and IV. These six exons give rise to five different
transcript variants, which encode three different protein iso-
forms (see Fig. 2 and Table 2). From the six exons present in
the mouse gene, only four (Ia, II, III, and IV) are conserved
in the human counterpart. The human gene does not con-
tain equivalents to mouse exons Ic and IIIb, and the mouse
gene lacks a region homologous to human exon Ib. Humans
and mice share the mitochondrial protein isoform Grx2a as
well as the second major protein isoform, the cytosolic/nu-
clear Grx2c. Remarkably, Grx2c is expressed by different
mechanisms in both species. In humans, it is derived from
alternative splicing of exon Ib and therefore is restricted to
testes (22). In mice, it is derived from alternative splicing of
exons Ia and Ic. Cytosolic Grx2 is present in distinct cells of
several mouse tissues, whereas neighboring cell types ex-
press primarily mitochondrial Grx2. Jurado et al. (17) ana-
lyzed the absolute gene-expression pattern of mouse Grx2
before. The number of mRNA molecules containing a frag-
ment spanning from the 3? end of exon I (Ia) to exon II cor-
responded to only 40% of the total number of Grx2 mRNA
molecules. As the exon Ia primer binding site in (17) was lo-
cated 3? of the alternative splice donor site in exon Ia, we can
conclude that at least part of the “missing” mRNA corre-
sponded to mGLRX2_v2 encoding Grx2c.
This widely expressed cytosolic isoform of Grx2 in mouse
was the most surprising finding of this study. What could
be the function of a second Grx in the cytosol of some spe-
cific cells? Grx1 and Grx2 both catalyze the reduction of
GSH-mixed disulfides with high specificity (16); however, in
vitro Grx2 can be activated by oxidation, whereas the Grx1
activity is inhibited when the additional structural cysteine
residues become oxidatively modified (10, 20). One might
speculate that Grx2c serves as backup for Grx1 under cer-
tain conditions. Recently, two published studies described
the effects of targeted disruption of the Grx1 gene in mice
(11, 26). Surprisingly, lack of function of Grx1 was not lethal
for mice, as described for Trx1 and Trx2 (29, 33). Moreover,
the first study reported that loss of Grx1 did not sensitize
adult mice to ischemia/reperfusion-induced injury or hy-
poxia, although embryonic fibroblasts were sensitized to ox-
idative stress (11). Somewhat controversy, the second study
described that Grx1 deficiency depressed functional recov-
ery and increased infarct size in coronary occlusion/reper-
fusion models of heart infarction and increased ROS pro-
duction during ischemia and reperfusion (26). It is tempting
to speculate that the mild phenotype described for the Grx1
knockout animals may be the result of complementation by
transcriptional activation of Grx2c. In contrast to this hy-
pothesis, Hoet al. (11) could not detect any compensatory de-
glutathionylation activity in different tissues of mice lacking
Grx1. Of course, this does not exclude that cytosolic Grx2
may compensate for other important activities of Grx1 (for
instance, the donation of electrons to RNR). It will be inter-
esting to see the expression and distribution of cytosolic Grx2
in these animals.
Our results confirm earlier observations (17) that mice
testes lack significant amounts of mGLRX2_v1 encoding
Grx2a. Instead, our data demonstrate the transcription of
HUDEMANN ET AL.10
FIG. 7.
croscopy. Co-staining with the cytosolic maker DNase I that stains globular actin, and the two mitochondrial marker pro-
teins prohibitin 1 and manganese superoxide dismutase (Mn-SOD) was described in detail in the experimental procedures
section. (A,B) Co-staining of Grx2 in spermatogonia with DNase I (A) but not mitochondrial prohibitin 1 (B). (C, D) In en-
teroendocrine cells of the fundic gland in stomach, Grx2 colocalizes with neither mitochondrial Mn-SOD (C) nor with pro-
hibitin 1 (D). (E, F) Cells of the red pulpa of the spleen lack co-staining of Grx2 with DNase I (E), but show a high degree
of colocalization with mitochondrial prohibitin 1 (F). (G, H) Cells of the white pulpa of spleen, conversely, show a high de-
gree of overlap with the cytosolic marker (G), but only a small degree of overlap with the mitochondrial marker (H).
Immunological detection of cytosolic Grx2 in cells of the testis, the stomach, and the spleen with confocal mi-

Page 11

FIG. 7.

Page 12

three additional mRNA variants, mGLRX2_v3, v4, and v5
encoding Grx2d and Grx2c, respectively, in testis. Immuno-
histochemical staining of Grx2 in the cytosol of spermato-
gonia provides evidence for the presence of Grx2c in these
cells. The transcript of variant 3 demonstrated by both RT-
PCR and 5?RACE is the only evidence so far for the putative
protein isoform Grx2d. It may be worth mentioning that exon
IIIb appears to be unique to mice. It is absent from human
and contains five nucleotide deletions over three different
parts in the rat genome. We can estimate from the relative
intensities of the bands shown in Fig. 1 that v4 (Grx2c) ac-
counts for ?50%, and both v3 (Grx2d) and v5 (Grx2c), for
about 25% of the total Grx2 mRNA in testicular tissue. What
could be the function of additional Grx2 isoforms in testis?
Spermatogenesis is a highly complex process of cellular dif-
ferentiation that takes place in the seminiferous tubules of
the testis. The male germ cell differentiation starts with the
proliferation of diploid germ cells (spermatogonia) followed
by meiotic division of spermatocytes, and finally, differenti-
ation of haploid spermatides into sperm. Cell division and
DNA synthesis require a constant supply of deoxyribonu-
cleotides by RNR. Surprisingly, neither of the established
electron donors for RNR (Trx1 and Grx1) colocalizes with
the enzyme in rat and calf testes. RNR is localized primar-
ily in spermatogonia (9, 36); Trx1 was found to be highly ex-
pressed in the interstitial Leydig cells and a small fraction of
spermatogonia (9, 37). Prominent staining of Grx1 was de-
tected in Sertoli cells, and weak staining, in Leydig cells. We
showed that Grx2c can serve as electron donor for RNR in
sub-micromolar concentrations (i.e., physiologic concentra-
tions). Thus, cytosolic Grx2 is the only confirmed electron
donor for RNR that colocalizes with the enzyme in sper-
matogonia.
During maturation of spermatids and sperm, oxidative
crosslinking of proteins through specific disulfide linkages
leads to the formation of structural sperm components (42,
47). A number of members of the Trx family of proteins
take part in these events: spermatocyte/spermatid-specific
thioredoxin 1 and 2, which both contain a Trx domain as
part of a larger protein, are found in the fibrous sheet of
the sperm (31, 39); spermatocyte/spermatid-specific thiore-
doxin 3 is associated with the Golgi apparatus of sper-
matids (14); and thioredoxin like 2 is associated with mi-
crotubuli in cilia and flagella (40). TGR staining was
detected in elongating spermatids and associated with the
vicinities of the assembling mitochondrial sheath (44). Sim-
ilar to Grx2, the phospholipid hydroperoxide reductase
(PHGPx) gene gives rise to two testis-specific isoforms tar-
geted to mitochondria and the nucleus, respectively (18, 34,
35, 41). PHGPx is present in soluble form in early sper-
matids but persists in mature sperm as a structural com-
ponent of the mitochondrial midpiece capsule (48). The ap-
parent colocalization of Grx2 with acrosomal structures in
both round and elongated spermatids suggests a similar
structural function of the protein or a role in the formation
of these disulfides.
In conclusion, we detected and characterized two new
Grx2 protein isoforms in mice: cytosolic Grx2c, which can
complex with a 2Fe2S cluster and serve as electron donor for
RNR, and a testis-specific variant, Grx2d, of unknown func-
tion. These results suggest additional specific functions of
Grx2 in multiple tissues of mice.
Acknowledgments
Christoph Hudemann and Maria Elisabet Lönn contrib-
uted equally to this work. We thank Drs. Elias Arnér and
Hans-Peter Elsässer for helpful discussions, Waltraud Ack-
ermann and Sabrina Oesteritz for excellent technical assis-
tance. and Karin Beimborn, Gisela Lesch. and Lena Ringdén
for excellent administrative assistance. This work was sup-
ported by grants from the Deutsche Forschungsgemeinschaft
(SFB593-N01), Karolinska Institutet, the Swedish Cancer So-
ciety, and the Swedish Society for Medical Research. F.Z.A.
was supported by a fellowship by the ministry of Health I.R.
Iran.
Abbreviations
ELISA, enzyme-linked immunosorbent assay; EST, ex-
pressed sequence tag; G3PDH, glyceralaldehyde-3-phos-
phate dehydrogenase; Grx, glutaredoxin; GSH, glutathione
(reduced); GSSG, glutathione disulfide; HED, hydroxyethyl
disulfide; Mn-SOD, manganese superoxide dismutase; RNR,
ribonucleotide reductase; ROS, reactive oxygen species; RT-
PCR, reverse transcription polymerase chain reaction; TGR,
thioredoxin glutathione reductase; Trx, thioredoxin; TrxR,
thioredoxin reductase; UTR, untranslated region.
References
1. Altschul SF, Gish W, Miller W, Meyers EW, and Lipman DJ.
Basic local alignment search tool. J Mol Biol 215: 403–410,
1990.
2. Beer SM, Taylor ER, Brown SE, Dahm CC, Costa NJ,
Runswick MJ, and Murphy MP. Glutaredoxin 2 catalyses the
reversible oxidation and glutathionylation of mitochondrial
membrane thiol proteins. J Biol Chem 279: 47939–47951, 2004.
3. Berndt C, Hudemann C, Hanschmann EM, Axelsson R,
Holmgren A, and Lillig CH. How does iron-sulfur cluster
coordination regulate the activity of human glutaredoxin 2?
Antioxid Redox Signal 9: 151–157, 2007.
4. Berndt C, Lillig CH, and Holmgren A. Thiol-based mecha-
nisms of the thioredoxin and glutaredoxin systems: impli-
cations for diseases in the cardiovascular system. Am J Phys-
iol Heart Circ Physiol 292: 1227–1236, 2007.
5. Chrestensen CA, Starke DW, and Mieyal JJ. Acute cadmium
exposure inactivates thioltransferase (glutaredoxin), inhibits
intracellular reduction of protein-glutathionyl-mixed disul-
fides, and initiates apoptosis. J Biol Chem 275: 26556–26565,
2000.
6. Davis R, Thelander M, Mann GJ, Behravan G, Soucy F,
Beaulieu P, Lavallee P, Graslund A, and Thelander L. Pu-
rification, characterization, and localization of subunit in-
teraction area of recombinant mouse ribonucleotide reduc-
tase R1 subunit. J Biol Chem 269: 23171–23176, 1994.
7. Enoksson M, Fernandes AP, Prast S, Lillig CH, Holmgren
A, and Orrenius S. Overexpression of glutaredoxin 2 atten-
uates apoptosis by preventing cytochrome c release. Biochem
Biophys Res Commun 327: 774–779, 2005.
8. Gladyshev VN, Liu A, Novoselov SV, Krysan K, Sun QA,
Kryukov VM, Kryukov GV, and Lou MF. Identification and
characterization of a new mammalian glutaredoxin (thi-
oltransferase), Grx2. J Biol Chem 276: 30374–30380, 2001.
9. Hansson HA, Rozell B, Stemme S, Engstrîm Y, Thelander L,
and Holmgren A. Different cellular distribution of thiore-
doxin and subunit M1 of ribonucleotide reductase in rat tis-
sues. Exp Cell Res 163: 363–369, 1986.
HUDEMANN ET AL. 12

Page 13

10. Hashemy SI, Johansson C, Berndt C, Lillig CH, and Holm-
gren A. Oxidation and S-nitrosylation of cysteines in human
cytosolic and mitochondrial glutaredoxins: effects on struc-
ture and activity. J Biol Chem 282:14428–14436, 2007.
11. Ho YS, Xiong Y, Ho DS, Gao J, Chua BH, Pai H, and Mieyal
JJ. Targeted disruption of the glutaredoxin 1 gene does not
sensitize adult mice to tissue injury induced by ischemia/
reperfusion and hyperoxia. Free Radic Biol Med 43: 1299–
1312, 2007.
12. Holmgren A. Hydrogen donor system for Escherichia coli
ribonucleoside-diphosphate reductase dependent upon glu-
tathione. Proc Natl Acad Sci U S A 73: 2275–2279, 1976.
13. Holmgren A. Thioredoxin and glutaredoxin systems. J Biol
Chem 264: 13963–13966, 1989.
14. Jimenez A, Zu W, Rawe VY, Pelto-Huikko M, Flickinger CJ,
Sutovsky P, Gustafsson JA, Oko R, and Miranda-Vizuete A.
Spermatocyte/spermatid-specific thioredoxin-3, a novel
Golgi apparatus-associated thioredoxin, is a specific marker
of aberrant spermatogenesis. J Biol Chem 279: 34971–34978,
2004.
15. Johansson C, Kavanagh KL, Gileadi O, and Oppermann U.
Reversible sequestration of active site cysteines in a 2FE2S-
bridged dimer provides a mechanism for glutaredoxin 2 reg-
ulation in human mitochondria. J Biol Chem 282: 3077–3082,
2007.
16. Johansson C, Lillig CH, and Holmgren A. Human mito-
chondrial glutaredoxin reduces S-glutathionylated proteins
with high affinity accepting electrons from either glu-
tathione or thioredoxin reductase. J Biol Chem 279:
7537–7543, 2004.
17. Jurado J, Prieto-Alamo MJ, Madrid-Risquez J, and Pueyo C.
Absolute gene expression patterns of thioredoxin and
glutaredoxin redox systems in mouse. J Biol Chem 278:
45546–45554, 2003.
18. Kelner MJ and Montoya MA. Structural organization of
the human selenium-dependent phospholipid hydroper-
oxide glutathione peroxidase gene (GPX4): chromosomal
localization to 19p133. Biochem Biophys Res Commun 249:
53–55, 1998.
19. Lillig CH and Holmgren A. Thioredoxin and related mole-
cules: from biology to health and disease. Antioxid Redox Sig-
nal 9: 25–47, 2007.
20. Lillig CH, Berndt C, Vergnolle O, Lönn ME, Hudemann C,
Bill E, and Holmgren A. Characterization of human glutare-
doxin 2 as iron-sulfur protein: a possible role as redox sen-
sor. Proc Natl Acad Sci U S A 102: 8168–8173, 2005.
21. Lillig CH, Lönn ME, Enoksson M, Fernandes AP, and Holm-
gren A. Short interfering RNA-mediated silencing of
glutaredoxin 2 increases the sensitivity of HeLa cells to-
wards doxorubicin and phenylarsine oxide. Proc Natl Acad
Sci U S A 101: 13227–13232, 2004.
22. Lönn ME, Hudemann C, Berndt C, Cherkasov V, Capani F,
Holmgren A, and Lillig CH. Expression pattern of human
glutaredoxin 2 isoforms: identification and characterization
of two testis/cancer cell-specific isoforms. Antioxid Redox
Signal 10: 547–558, 2008.
23. Lundberg M, Fernandes AP, Kumar S, and Holmgren A.
Cellular and plasma levels of human glutaredoxin 1 and 2
detected by sensitive ELISA systems. Biochem Biophys Res
Commun 319: 801–809, 2004.
24. Lundberg M, Johansson C, Chandra J, Enoksson M, Jacobs-
son G, Ljung J, Johansson M, and Holmgren A. Cloning and
expression of a novel human glutaredoxin (GRX2) with mi-
tochondrial and nuclear isoforms. J Biol Chem 276:
26269–26275, 2001.
25. Luthman M and Holmgren A. Glutaredoxin from calf thy-
mus purification to homogeneity. J Biol Chem 257: 6686–6689,
1982.
26. Malik G, Nagy N, Ho YS, Maulik N, and Das DK. Role of
glutaredoxin-1 in cardioprotection: an insight with Glrx1
transgenic and knockout animals. J Mol Cell Cardiol 44:
261–269, 2008.
27. Mann GJ, Graslund A, Ochiai E, Ingemarson R, and The-
lander L. Purification and characterization of recombinant
mouse and herpes simplex virus ribonucleotide reductase
R2 subunit. Biochemistry 30: 1939–1947, 1991.
28. Martin JL. Thioredoxin: a fold for all reasons. Structure 3:
245–250, 1995.
29. Matsui M, Oshima M, Oshima H, Takaku K, Maruyama T,
Yodoi J, and Taketo MM. Early embryonic lethality caused
by targeted disruption of the mouse thioredoxin gene. Dev
Biol 178: 179–185, 1996.
30. Miranda-Vizuete A, Damdimopoulos AE, and Spyrou G.
The mitochondrial thioredoxin system. Antioxid Redox Sig-
nal 101: 13227–13232, 2000.
31. Miranda-Vizuete A, Ljung J, Damdimopoulos AE, Gustafs-
son JA, Oko R, Pelto-Huikko M, and Spyrou G. Characteri-
zation of Sptrx, a novel member of the thioredoxin family
specifically expressed in human spermatozoa. J Biol Chem
276: 31567–31574, 2001.
32. Murata H, Ihara Y, Nakamura H, Yodoi J, Sumikawa K, and
Kondo T. Glutaredoxin exerts an antiapoptotic effect by reg-
ulating the redox state of Akt. J Biol Chem 278: 50226–50233,
2003.
33. Nonn L, Williams RR, Erickson RP, and Powis G. The ab-
sence of mitochondrial thioredoxin 2 causes massive apop-
tosis, exencephaly, and early embryonic lethality in ho-
mozygous mice. Mol Cell Biol 23: 916–922, 2003.
34. Pfeifer H, Conrad M, Roethlein D, Kyriakopouluos A,
Brielmeier M, Bornkamm GW, and Behne D. Identification
of a specific sperm nuclei selenoenzyme necessary for pro-
tamine thiol cross-linking during sperm maturation. FASEB
J 15: 1236–1238, 2001.
35. Roveri A, Maiorino M, Nissi C, and Ursini F. Purification
and characterization of phospholipid hydroperoxide glu-
tathione peroxidase from rat testis mitochondrial mem-
branes. Biochim Biophys Acta 1208: 211–221, 1994.
36. Rozell B, Barcena JA, Martinez-Galisteo E, Padilla CA, and
Holmgren A. Immunochemical characterization and tissue
distribution of glutaredoxin (thioltransferase) from calf. Eur
J Cell Biol 62: 314–323, 1993.
37. Rozell B, Hansson HA, Luthman M, and Holmgren A. Im-
munohistochemical localization of thioredoxin and thiore-
doxin reductase in adult rats. Eur J Cell Biol 38: 79–86, 1985.
38. Rudolph R and Lilie H. In vitro folding of inclusion body
proteins. FASEB J 10: 49–56, 1996.
39. Sadek CM, Damdimopoulos AE, Pelto-Huikko M, Gustafs-
son JA, Spyrou G, and Miranda-Vizuete A. Sptrx-2, a fusion
protein composed of one thioredoxin and three tandemly re-
peated NDP-kinase domains is expressed in human testis
germ cells. Genes Cells 6: 1077–1091, 2001.
40. Sadek CM, Jimenez, A, Damdimopoulos AE, Kieselbach T,
Nord M, Gustafsson JA, Spyrou G, Davies EC, Oko R, van
der Hoorn FA, and Miranda-Vizuete A. Characterization of
human thioredoxin-like 2 A novel microtubule-binding
thioredoxin expressed predominantly in the cilia of lung air-
way epithelium and spermatid manchette and exoneme. J
Biol Chem 278: 13133–13142, 2003.
41. Schneider M, Vogt Weisenhorn DM, Seiler A, Bornkamm
GW, Brielmeier M, and Conrad M. Embryonic expression
ISOFORMS OF MOUSE Grx2 13

Page 14

profile of phospholipid hydroperoxide glutathione peroxi-
dase. Gene Expr Patterns 6: 489–494, 2006.
42. Shalgi R, Seligman J, and Kosower NS. Dynamics of the thiol
status of rat spermatozoa during maturation: analysis with
the fluorescent labeling agent monobromobimane. Biol Re-
prod 40: 1037–1045, 1989.
43. Shelton MD, Chock PB, and Mieyal JJ. Glutaredoxin: role in
reversible protein S-glutathionylation and regulation of re-
dox signal transduction and protein translocation. Antioxid
Redox Signal 7: 348–366, 2005.
44. Su D, Novoselov SV, Sun QA, Moustafa ME, Zhou Y, Oko
R, Hatfield DL, and Gladyshev VN. Mammalian selenopro-
tein thioredoxin-glutathione reductase: roles in disulfide
bond formation and sperm maturation. J Biol Chem 280:
26491–26498, 2005.
45. Sun C, Berardi MJ, and Bushweller JH. The NMR solution
structure of human glutaredoxin in the fully reduced form.
J Mol Biol 280: 687–701, 1998.
46. Sun QA, Kirnarskydagger L, Shermandagger S, and Glady-
shev VN. Selenoprotein oxidoreductase with specificity for
thioredoxin and glutathione systems. Proc Natl Acad Sci U S
A 98: 3673–3678, 2001.
47. Sutovsky P, Tengowski MW, Navara CS, Zoran SS, and Schat-
ten G. Mitochondrial sheath movement and detachment in
mammalian, but not nonmammalian, sperm induced by disul-
fide bond reduction. Mol Reprod Dev 47: 79–86, 1997.
48. Ursini F, Heim S, Kiess M, Maiorino M, Roveri A, Wissling
J, and Flohe L. Dual function of the selenoprotein PHGPx
during sperm maturation. Science 285: 1393–1396, 1999.
49. Wang J, Boja ES, Tan W, Tekle E, Fales HM, English S, Mieyal
JJ, and Chock PB. Reversible glutathionylation regulates
actin polymerization in A431 cells. J Biol Chem 276:
47763–47766, 2001.
50. Wells WW, Xu DP, Yang YF, and Rocque PA. Mammalian
thioltransferase (glutaredoxin) and protein disulfide iso-
merase have dehydroascorbate reductase activity. J Biol
Chem 265: 15361–15364, 1990.
Address reprint requests to:
Christopher Horst Lillig
Institute for Clinical Cytobiology and Cytopathology
Phillips UniversitÑt
DE-35037 Marburg, Germany
E-mail: horst@lillig.de
Date of first submission to ARS Central, March 5, 2008;
date of final revised submission, June 24, 2008; date of
acceptance, June 24, 2008.
HUDEMANN ET AL. 14