Cloning, Sequencing, and Characterization of Alternatively Spliced Glutaredoxin 1 cDNA and Its Genomic Gene

Alternatively spliced human glutaredoxin (Grx1as) cDNA was isolated from a neutrophil cDNA library, using a 32P-labeled human glutaredoxin (Grx1) cDNA probe under non-stringent conditions. The sequence of Grx1as cDNA indicated that the open reading frame of the gene was identical to the open reading frame of the previously reported first human glutaredoxin (Grx1) cDNA, but the 3′-untranslated region of Grx1as was not homologous to Grx1 cDNA. Northern blot and RT-PCR analyses showed Grx1as mRNA was expressed in normal human neutrophils and transformed cells including U937, HL-60, THP, and Jurkat cells. Cloning and sequencing of the genomic gene corresponding to Grx1as cDNA showed that two different glutaredoxin cDNAs (Grx1as and Grx1) were generated from the same genomic gene via alternative splicing. Origination of Grx1as and Grx1 from the same gene was confirmed by chromosomal localization of the Grx1as gene to chromosome 5q13, the same location where the Grx1 gene was localized previously. During screening of the Grx1as genomic gene, two additional glutaredoxin pseudogenes were also isolated. Surprisingly, these pseudogenes contained 3′-untranslated regions that were nearly identical to the 3′-untranslated regions of Grx1as, not Grx1, cDNA. Because 3′-untranslated regions may be important in stabilizing mRNAs, the effect of the two 3′-untranslated regions of Grx1 and Grx1as on mRNA stability was investigated using luciferase reporter vectors with the 3′-untranslated regions. Luciferase activity was 2.6-fold greater in cells transfected with the reporter vector containing the 3′-untranslated region of Grx1as cDNA compared with the 3′-untranslated region of Grx1 cDNA. These data indicate that Grx1as cDNA is an alternatively spliced human Grx1 cDNA and that the Grx1as 3′-untranslated region may have a role in stabilizing mRNA.

Glutaredoxin (Grx, 1 thioltransferase) is a small redox protein involved in oxidoreductive processes in cells through cat-alyzing disulfide-thiol exchange reactions (1)(2)(3). Thiol groups in proteins may act as redox-sensitive switches and are considered to be a key element in maintaining cellular redox balance (4 -8). Cellular redox imbalance induces radical oxygen species that mediate signaling events leading to proliferation, apoptosis, and differentiation (4,5). Because of its potential importance, the redox status of thiol groups is well balanced by biological reducing molecules and proteins (6 -10). Among redox proteins, glutaredoxin is a protein that regenerates Sthiolated cysteines in proteins that result from oxidative stress (11). Glutaredoxins have been isolated from prokaryotes and eukaryotes and have been proposed as redox proteins that mediate several biological reactions (3,(11)(12)(13)(14)(15)(16). Recently, a second human glutaredoxin (glutaredoxin 2 or Grx2) has been cloned, and two alternatively spliced Grx2 mRNA isoforms were identified (17).
In our laboratory, we searched for mRNA isoforms of the first human glutaredoxin (glutaredoxin 1 or Grx1), because swine glutaredoxin cDNA contains a 3Ј-untranslated region that has no counterpart in human glutaredoxin 1 cDNAs (18,19). A human neutrophil library was screened with 32 P-labeled Grx1 cDNA using non-stringent conditions. Screening yielded a new human glutaredoxin cDNA, Grx1 as . Its nucleotide sequence was identical to the open reading frame of Grx1 cDNA, but the 3Ј-untranslated region sequence was comparable to that of swine glutaredoxin cDNA rather than that of human Grx1 cDNA. Northern blots and RT-PCR were performed to verify Grx1 as mRNA in several human cells. Genomic cloning and chromosomal localization of Grx1 as were performed to determine genomic origins of Grx1 and Grx1 as cDNAs. Finally effects of 3Ј-untranslated regions from Grx1 as and Grx1 cDNAs on mRNA stability were investigated.
Construction of Human Neutrophil cDNA and Genomic Library-A human neutrophil cDNA library was constructed as described previously (3). Briefly, first and second strand cDNA were synthesized (Stratagene, La Jolla, CA) using 5 g of human neutrophil mRNA obtained from 10 subjects, each of whom provided ϳ1 ϫ 10 9 neutrophils isolated by apheresis (20,21). The cDNAs were ligated to Uni-ZAP * This work was supported in part by Grant Z01 DK 54506 from NIDDK, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF069668, NM002064, AF115104, AF115105, AF115106, HSU61726, and AY918930.
Isolation and Sequencing of Alternatively Spliced Human Glutaredoxin (Grx1 as ) cDNA-An amplified human neutrophil cDNA library was screened with random-primed 32 P-labeled partial human neutrophil glutaredoxin cDNA from nucleotides 50 to 270 of the open reading frame (3). Several clones were selected from 2 ϫ 10 6 recombinants, and clones were further purified by sequential platings. The size and integrity of cDNA was determined by PCR with primers corresponding to various regions of the human cDNA. Nucleotide sequences were determined using the dideoxy chain termination method with modified T7 DNA polymerase (Amersham Biosciences).
Northern Blot of Grx1 as cDNA-Northern blot was performed using 2 g of poly(A) ϩ RNAs isolated from neutrophils, U937, HL-60, PLB, and Jurkat cells. The blot was hybridized under stringent conditions (22) with a probe of the 3Ј-untranslated region of Grx1 as cDNA.
Isolation and Sequencing of Genomic Grx1 as Gene-An amplified human neutrophil genomic DNA library was screened with randomprimed 32 P-labeled Grx1 as 3Ј-untranslated region cDNA. Two overlapping positive clones were selected from 8 ϫ 10 6 recombinants. The plaques were purified and designated G1-G2. The size and integrity of insert genomic DNA was determined by PCR with primers corresponding to various regions of the human cDNA. PCR reactions were performed with a DNA Thermal Cycler according to the manufacturer's recommendations (PerkinElmer Life Sciences). One (G1) of two clones contained the entire glutaredoxin gene, the 1.2-kb upstream 5Ј-flanking region and the 3-kb downstream 3Ј-flanking region. A genomic fragment of ϳ10 kb in clone G1 was analyzed with PCR. The DNA containing the glutaredoxin promoter region was amplified using the 18 oligomer (5Ј-CACAAACTCTTGAGCCAT; primer 1) of the antisense strand complementary to nucleotide positions 1-18 of glutaredoxin cDNA, and T3 primer located in the left arm. Glutaredoxin genomic gene was digested with SacI, and the sequences of digested DNA fragments were determined using the dideoxy chain termination method with modified T7 DNA polymerase and the ALF DNA sequencer (Amersham Biosciences). During the genomic screening, we also isolated two glutaredoxin pseudogenes (GS1 and GS2).
Chromosomal Localization-The Grx1 as gene was labeled with digoxigenin dUTP by nick translation. Labeled probe was combined with sheared human DNA and hybridized to normal metaphase chromosomes derived from phytohemagglutinin-stimulated peripheral blood lymphocytes in a solution containing 50% formamide, 10% dextran sulfate, and 2ϫ sodium chloride, sodium citrate buffer (23).
Determination of Effect of 3Ј-Untranslated Regions on mRNA Stability-To measure the effect of short and long 3Ј-untranslated regions on mRNA stability, a luciferase reporter vector (pCMV-Luc) was constructed by cloning PCR-amplified luciferase gene into pCR 3.1 (Invitrogen). The reporter vectors contained either the short or long 3Ј-untranslated region of human glutaredoxin: pCMV-Luc-3Ј-untranslated region Grx1 or pCMV-Luc-3Ј-untranslated region Grx1 as , respectively. The vectors were constructed by insertion of each 3Ј-untranslated region downstream of the luciferase gene using the PstI site. Vectors were transfected with calcium phosphate into HeLa cells. Cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. Cells were grown to ϳ50% confluence in 60-mm Petri dishes. The two reporter plasmids were individually transfected into cells using calcium phosphate. After transfection, cells were incubated for an additional 48 h, harvested, and lysed. Luciferase assays of the lysates were performed according to the manufacturer's protocol (Promega). Transfection efficiency was monitored by co-transfection with pSEAP-2 promoter vector, the alkaline phosphatase activity of which was determined according to the manufacturer's protocol (Clontech).
Isolation of Grx1 as and Grx1 Recombinant Proteins-Grx1 was amplified and purified as described (3). Grx1 as cDNA was amplified by PCR using two primers: 5Ј-ATGGCTCAAGAGTTTGTG-3Ј and 5Ј-TTACTGCAGAGCTCCAAT-3Ј, which were complementary to nucleotide positions 1-18 (sense) and 303-321 (antisense) of glutaredoxin cDNA, respectively. For expression, PCR fragments were cloned into pGEX expression vector (Amersham Biosciences). The vector was transformed into E. coli (BL21), and bacteria were cultured overnight at 37°C in Luria Bertani medium containing ampicillin (100 g/ml). Overnight culture was inoculated into fresh medium and cultured further with vigorous shaking. When OD 600 nm was 0.6, protein expression was induced by adding 0.5 mM isopropyl ␤-D-thiogalactopyranoside. After 4 h of induction, bacteria were collected by centrifugation. Recombinant   FIG. 1. The nucleotide sequence and deduced amino acid sequence of Grx1 as glutaredoxin cDNA. Sequences 1 and 2 represent Grx1 as and Grx1, respectively. The nucleotides are numbered 5Ј to 3Ј, and amino acids are numbered from N to C termini, respectively. Two conserved regions containing a total of four cysteines are underlined. Bold nucleotides indicates that they are the same in both clones, and *** indicates the stop codon.
glutathione was expressed as glutathione S-transferase fusion protein, and the fusion protein was purified using a glutathione-Sepharose column (Amersham Biosciences).
Dehydroascorbic Acid Reducing Activity-Reducing activity was determined in 25 l of 100 mM Tris-HCl buffer (pH 7.5) containing (final concentrations) 0.8 mM reduced glutathione, 500 M of dehydroascorbic acid, and appropriate amounts of enzyme. The reaction was initiated by adding dehydroascorbic acid and measured for 3 min at room temperature. The reaction was terminated by addition of 35 l of 90% methanol in water containing 1 mM EDTA. The final mixture was centrifuged for 10 min at 14,000 ϫ g, and the supernatant immediately analyzed by high performance liquid chromatography (24). A chemical reaction containing no enzyme was simultaneously measured under the same conditions. Where indicated chemical activity was subtracted from total reducing activity to yield enzymatic activity. Chemical activity was always measured and accounted for (3). The amount of reduced dehydroascorbic acid was determined as ascorbic acid by high performance liquid chromatography with coulometric electrochemical detection as described previously (24). Dehydroascorbic acid was always prepared fresh immediately prior to experiments as described (25).

RESULTS
Isolation of Grx1 as cDNA-To isolate different types of human glutaredoxin cDNA, partial cDNA glutaredoxin from nucleotides 50 to 270 of the open reading frame was used as a probe (3). This region was selected because it contained four conserved cysteines including an active site cysteine. Plaques (8 ϫ 10 6 ) were screened under low stringent conditions with the 32 P-labeled DNA probe. Positive plaques were selected, and the integrity of each plaque was tested by PCR, using internal primers located in the previously reported glutaredoxin cDNA (3, 27, 28) (data not shown). The nucleotide sequence of the isolated clone was determined and compared with that of Grx1 cDNA (Fig. 1). Surprisingly, the nucleotide sequences of the translation regions of both cDNAs were homologous, but the 3Ј-untranslated region of the newly isolated cDNA was different from that of Grx1 cDNA. To date, at least three mammalian glutaredoxin cDNAs have been reported, among which is swine glutaredoxin cDNA. Swine glutaredoxin cDNA contains a translation region exhibiting ϳ80% identity to the human counterpart, but a region with no similarity is found in the 3Ј-untranslated regions of human and swine glutaredoxin cDNAs (19). The newly isolated glutaredoxin cDNA not only exhibited an identical nucleotide sequence to that of previously reported human glutaredoxin (Grx1) cDNA in the translation region, but also was quite similar to the 3Ј-untranslated region counterpart of swine glutaredoxin cDNA rather than Grx1 cDNA. We termed the newly isolated glutaredoxin cDNA Grx1 as .
Grx1 as Is Present in Multiple Cell Types-Expression of Grx1 as mRNA was determined by Northern blot and RT-PCR with mRNAs from human neutrophils, U937, HL-60, THP, and Jurkat cells. Transfers of mRNAs onto nitrocellulose membrane were performed using a standard method (22). The probe used was not a full sequence of Grx1 as cDNA, because the translation region of Grx1 as cDNA contained the same nucleotide sequence as Grx1 cDNA. The probe was amplified by PCR with the two primers (forward and reverse primers located at nucleotides 350 -370 and 520 -540, respectively in Fig. 1). The  amplified DNA fragment that had no homology to Grx1 cDNA was 32 P-labeled by the random primer method. The transferred nitrocellulose membrane was hybridized using stringent conditions with the 32 P-labeled probe. Grx1 as cDNA was expressed in all the cells tested (Fig. 2). The data indicate that Grx1 as is expressed in cells of myelocytic and lymphocytic origins.
There are several indications that Grx1 as is a generally observed phenomenon and not an artifact from a single subject source. The library from which Grx1 as was originally isolated was prepared from neutrophils from 10 subjects. As shown in Fig. 2, Grx1 as expression was verified by Northern blot in all cells tested of myeloid origin: HL60 cells, THP cells, Jurkat cells, and PLB cells. As confirmation of these data, RT-PCR with specific primers for Grx1 as showed that the mRNA was present in all the myeloid cell types tested (data not shown). Finally, when Grx1 as specific primers were used, the mRNA was also present in a commercial myeloid library.
Characterization of Recombinant Grx1 as -The nucleotide sequences of Grx1 as and Grx1 cDNAs are identical in their translation regions. Therefore, we predicted that the recombinant protein from Grx1 as cDNA would be likely to exhibit the same biochemical and kinetics properties as glutaredoxin from Grx1 cDNA. To study this, the translation region (open reading frame) of Grx1 as cDNA was amplified and ligated into pGEX fusion expression vector as described under "Experimental Procedures." Glutathione S-transferase-fused protein was purified using a glutathione-Sepharose column, and the fusion protein was digested with thrombin (0.1%, w/w) to yield recombinant protein. Glutaredoxin, from Grx1 cDNA, was isolated and purified as described (3).
Biochemical properties of the recombinant proteins from Grx1 and Grx1 as cDNAs were characterized with respect to dehydroascorbic acid reducing activities ( Table I). The following properties were virtually identical: apparent K m for glutathione; apparent K m for dehydroascorbic acid; optimal pH; and specific activity. Each recombinant purified protein was detected by Western blot (Fig. 3), using polyclonal antibody prepared by injecting Grx1 in rabbits (3). Taken together, these data indicate that biochemical and kinetics properties of the two proteins were indistinguishable.
Isolation and Sequencing of the Genomic Gene of Grx1 as -To delineate the origin of Grx1 as cDNA, its genomic gene was screened in a neutrophil genomic library with the same probe used in the Northern blot. A positive plaque was selected as the candidate for the genomic gene of Grx1 as cDNA. The integrity and authenticity of the clone were checked by PCR using several primers (see "Experimental Procedures"). To obtain detailed information on the nucleotide sequence of the isolated genomic gene, the clone was digested with restriction enzymes and subcloned into pGEM sequence vector. Surprisingly, the nucleotide sequence of the 5Ј-flanking region of Grx1 as cDNA was exactly the same as that of the published genomic gene of Grx1 cDNA (17,18). Due to this unexpected result, the nucleotides of other regions of the isolated genomic gene were further sequenced. We found that the sequenced regions of the Grx1 as genomic gene were perfectly matched to the corresponding region of the Grx1 genomic gene (17,18). Therefore, every individual subclone was amplified by PCR with specific primers used in the Northern blot in order to identify the subclone containing the unique 3Ј-untranslated region of Grx1 as cDNA. One subclone was confirmed by PCR to contain the 3Ј-untranslated region, and sequenced for its verification. For clarity, a partial nucleotide sequence of this subclone is shown in Fig. 4. The subclone contained, in order as shown: an overlapping partial sequence of an intron; the 3Ј-untranslated region of Grx1 as cDNA (as shown in Fig. 1); a unique 3Ј-untranslated region of Grx1 as cDNA; and the 3Ј-untranslated region of Grx1 cDNA (also as shown in Fig. 1). The full sequence of this subclone is shown in Fig. 5 and represents the nucleotide sequence of the genomic gene of Grx1 as . The sequence includes the coding region that is identical to glutaredoxin, and contains two introns and three exons. This gene organization is identical to that of the genomic gene for Grx1 (17,18) (Fig. 6). Taken together, these data indicate that Grx1 and Grx1 as cDNAs are produced from the same human glutaredoxin genomic gene via alternative splicing at gggcag/AACAGGCCC in the second in-tron instead of ccacag/ATCTCATAG (Figs. 5 and 6). Thus, Grx1 as cDNA, but not Grx1, includes an additional 566 nucleotides from second intron in its 3Ј-untranslated region. The alternative splicing of human glutaredoxin genomic gene explains why human Grx1 as and swine glutaredoxin cDNAs, but not Grx1 cDNA, are similar in their 3Ј-untranslated regions. To confirm this, a full-length Grx1 as cDNA was cloned and completely sequenced. Sequencing the human glutaredoxin genomic gene showed that an isolated full-length cDNA of Grx1 as contained the unique sequence of 3Ј-untranslated region of Grx1 as , and 3Ј-untranslated region of Grx1 (Fig. 7). The 3Ј-untranslated region of Grx1 as cDNA is approximately twice the size of the counterpart of the 3Ј-untranslated region of Grx1 cDNA.
Chromosomal Localization of the Grx1 as Gene-The sequence of the genes from Grx1 and Grx1 as cDNAs indicated that each individual human glutaredoxin cDNA was probably transcribed from the same gene. This was confirmed by chromosomal localization of the isolated Grx1 as gene. Specific hybridization signals were detected by incubating the hybridized slides in fluoresceinated antidioxigenin antibodies followed by counterstaining with 4Ј-6-diamidino-2-phenylindole dihydrochloride. The results showed a specific labeling of chromosome 5, based on size, morphology, and banding pattern. In additional experiments, a genomic probe (previously mapped to 5q32 and confirmed by co-hybridization with a probe from the cri du chat locus on chromosome arm 5p) was co-hybridized with the hybridized probe of the Grx1 as gene. In these experiments the middle and distal long arm of chromosome 5 were specifically labeled. Measurement of ten specifically hybridized chromosomes 5 demonstrated that the glutaredoxin gene is located at a position which is 39% of the distance from the centromere to the telomere of chromosome arm 5q, an area which corresponds to the boundary between bands 5q15 and 5q21 (Fig. 8). This chromosomal position was previously mapped to the human Grx1 gene (29). The nucleotide sequence and chromosomal mapping of the human Grx1 as gene indicate that Grx1 as cDNA is undoubtedly transcribed from the same gene that human Grx1 cDNA comes from.
Effect of Two 3Ј-Untranslated regions of Grx1 and Grx1 as on Their mRNA Stability-During the screening of the Grx1 genomic gene, two different putative pseudogenes (GS1 and GS2) were isolated and sequenced. They seemed to display typical characteristics of pseudogenes: no intron, poly(A) tailing, and sporadic mutations. However, some regions of the 3Ј-untranslated regions of these genes were not homologous to the 3Ј-untranslated region of Grx1 cDNA. Instead, Grx1 as cDNA and the two glutaredoxin pseudogenes were very similar in portions of their 3Ј-untranslated regions (Fig. 9). It was surprising that the pseudogenes had greater resemblance to FIG. 6. Physical map and comparison of an isolated gene of Grx1 as with Grx1 gene (17). Grx1 and Grx1 as are depicted in I and II, respectively. The sequences of the junction regions between exons and introns are shown in III. Two cDNAs of Grx1 (I) and Grx1 as (II) are compared side by side to illustrate that the 3Ј-untranslated regions are not identical to each other. Different junctions between exons and introns are labeled alphabetically.
Grx1 as than to Grx1 cDNA, because pseudogenes are commonly considered to originate from their authentic mRNA. Stability of mRNA may therefore be a contributing factor for generating pseudogenes. We investigated this possibility by measuring the effect of the 3Ј-untranslated regions of Grx1 and Grx1 as on the stability of their mRNAs. The effect of the two 3Ј-untranslated regions was determined by constructing the vectors pCMV-Luc-3Ј-untranslated region Grx1 and pCMV-Luc-3Ј-untranslated region Grx1 as , transfecting these vectors into cells, and measuring luciferase activity (see "Experimental Procedures"). As shown in Table II, the decreasing order of luciferase activities was pCMV-Luc-3Ј-untranslated region Grx1 as Ͼ pCMV-Luc-3Ј-untranslated region Grx1 Ͼ pCMV-Luc control. Since all three reporter vectors have the same CMV promoter, the promoter activities of the vectors and their initial mRNAs may be very similar. To verify this assumption, quantitative RT-PCR was performed with mRNAs from HeLa cells transfected with each reporter vector. Amounts of luciferase mRNA from the three transfected cells was very similar (data not shown). These data indicate that the quantity of luciferase mRNAs was similar in the cells transfected with three different reporter vectors, but mRNA containing 3Ј-untranslated region Grx1 as might be more stable than the other two mRNAs, thereby producing more luciferase. Because the 3Ј-untranslated region of Grx1 as mRNA may confer enhanced stability, Grx1 as mRNA may be a better template than Grx1 mRNA for generating its pseudogenes. DISCUSSION On comparing the sequences of mammalian glutaredoxin cDNAs, we found that swine glutaredoxin cDNA contains a 3Ј-untranslated region non-homologous to human Grx1 cDNA (18,19). The discrepancy between human and swine cDNAs suggested that human glutaredoxin cDNA might exist in more than one form. It seemed reasonable that the existence of different human glutaredoxin cDNAs and a non-homologous region in swine glutaredoxin cDNA should be addressed prior to investigating glutaredoxin regulation, because it is possible that each glutaredoxin cDNA may be regulated differently in human cells. As a way to search for different human glutaredoxin mRNA, a human neutrophil cDNA library was screened with a probe of Grx1 cDNA using non-stringent conditions. In the screening process, a different human glutaredoxin (Grx1 as ) cDNA was isolated. Surprisingly, the nucleotide sequence and chromosomal localization of the genomic gene of Grx1 as cDNA indicated that Grx1 as and Grx1 were transcriptional products derived from the same gene via alternative splicing. The isolation, sequencing, and localization of Grx1 as cDNA provide an answer for the lack of identity between the 3Ј-untranslated region of human glutaredoxin (Grx1) and swine glutaredoxin.
However, it was still not clear whether two different glutaredoxin cDNAs resulted from a single transcript with alternative splicing or from two different transcripts generated using two different promoter regions of the same gene. Primer extension and S1 mapping experiments answered this question (18). Human neutrophil mRNA exhibited only one band protected from S1 nuclease, even though neutrophils contain two different glutaredoxin mRNAs (18) (Fig. 2). These data suggest that one promoter is apparently used for expression of a pre-mRNA leading to two different human glutaredoxin mRNAs via alternative splicing. Determining how this process proceeds may advance understanding of the cellular response to oxidative stress. In our preliminary studies, a certain region of the promoter of human glutaredoxin gene was identified as a transcriptional regulatory region, which may bind to several potential transcriptional factors. Future investigation will provide detailed information regarding the regulation of human glutaredoxin via this transcriptional regulatory region.
When we began this study, it was unclear whether isoforms of human glutaredoxin existed. However, human glutaredoxin 2 (Grx2) has been isolated recently, and two mRNAs of Grx2 have been identified as alternative splicing products (17). In-terestingly, both Grx1 and Grx2 have two mRNAs generated via alternative splicing, respectively. During our screening of the human glutaredoxin 1 gene, two additional genes were isolated and sequenced. Each of these genes has characteristics of a pseudogene: non-existence of introns, possession of poly(A) tail, and sporadic mutations (30). Of interest, the isolated pseudogenes (GS1 and GS2) resemble Grx1 as cDNA rather than Grx1 cDNA. In other words, each pseudogene contains a 3Јuntranslated region similar to the 3Ј-untranslated region of human Grx1 as cDNA. Furthermore, a glutaredoxin pseudogene (GS1) has a mutation of the first conserved cysteine to phenylalanine. Further investigation of this pseudogene is now underway to determine whether the pseudogene is expressed in cells, and what biological function the pseudogene may deliver, if expressed. Although alternative splicing is a common mechanism used to generate two mRNAs in Grx1 and Grx2, respectively, the biological consequences of the 3Ј-untranslated regions of Grx2 are largely unknown currently.
In this report, two different human glutaredoxin 1 mRNAs were demonstrated to exist, and potential biological consequences of the two mRNAs were elucidated. Glutaredoxin pseudogenes contain a long 3Ј-untranslated region, similar to that of Grx1 as cDNA. According to current hypotheses, mRNAs are templates for producing pseudogenes, and pseudogenes contain a poly(A) tail, sporadic mutations, and no introns. If so, human Grx1 as cDNA with a long 3Ј-untranslated region was perhaps more stable than Grx1 cDNA with a short 3Ј-untranslated region. Greater stability would increase the likelihood of Grx1 as serving as a template for human glutaredoxin pseudogenes. This study showed that a luciferase gene with the longer 3Ј-untranslated region of Grx1 as had a more stable transcript FIG. 9. Nucleotide sequence comparison of Grx1 as cDNA with two glutaredoxin pseudogenes (GS1 and GS2). The deduced amino acid sequence of Grx1 as is depicted on the top line. Sequences 1, 2, and 3 represent glutaredoxin I as , glutaredoxin pseudogene I (GS1), and glutaredoxin pseudogene II (GS2), respectively. The 3Ј-untranslated region unique to Grx1 as is in bold. Shaded areas in the pseudogenes represent nucleotide identity to the 3Ј-untranslated region unique to Grx1 as . Two cysteine regions are underlined; the region indicating stop codon is indicated by ***, and the signal sequence for the poly(A) tail is double underlined.

TABLE II
Effect of 3Ј-untranslated regions of Grx1 and Grx1 ast on mRNA stability as measured by luciferase activity Each 3Ј-untranslated region was fused to luciferase as described under "Experimental Procedures," and pCMV-Luc was the control reporter vector. Each of the three vectors was transfected into HeLa cells, and luciferase activities were determined as described under "Experimental Procedures." PCMV-Luc control pCMV-Luc-3Ј-untranslated region Grx1 pCMV-Luc-3Ј-untranslated region Grx1 as Luciferase activity 150 Ϯ 70 648 Ϯ 150 1711 Ϯ 180 than the shorter 3Ј-untranslated region of Grx1, as indicated by greater luciferase activity in transfected cells. In summary, the data in this paper provided information about the expression of different human glutaredoxin 1 cDNAs and their mRNA stability, and also may provide insight about the origin and potential biological consequence of the 3Ј-untranslated region from Grx2.