Rundlöf et al.5’ variants of TXNRD1 transcripts

Evidence for intriguingly complex regulation of human TXNRD1 transcription

Anna-Klara Rundlöf†#, Magnus Janard†, Antonio Miranda-Vizuete§ and Elias S. J. Arnér†#

†Medical Nobel Institute for Biochemistry, Department of Medical Biochemistry and Biophysics, Karolinska Institute, SE-171 77, Stockholm, and §, Center for Biotechnology, Department of Biosciences at NOVUM, Karolinska Institutet, SE-141 57, Huddinge, Sweden

# Corresponding authors:

Tel: +46-8-524 869 83

Fax: +46-8-31 15 51

E-mail: or

Running Title: 5’ variants of TXNRD1 transcripts

This study was supported by grants from The Swedish Society for Medical Research (to A.-K.R), the Swedish Medical Society (to E.S.J.A.), The Karolinska Institute (to E.S.J.A. and A.M.-V.), the Swedish Cancer Society (Projects 3775 and 4056 to E.S.J.A.), the Swedish Medical Research Council (Projects 03P-14096-01A and 03X-14041-01A to A.M.-V.) and the Åke Wibergs Stiftelse (A.M.-V.).

Evidence for intriguingly complex regulation of human TXNRD1 transcription

Abstract

Thioredoxin reductase 1 (TrxR1, TXNRD1) is a ubiquitously expressed selenoprotein with many important redox regulatory functions. In this study, we have further characterized the recently identified core promoter region of human TXNRD1. One critical Sp1/Sp3 site was found to be important in A549 and HeLa cells, whereas another Sp1/Sp3 site and one Oct-1 site bound transcription factors but were nonetheless dispensable for transcription. We also experimentally identified several 5'-region TXNRD1 transcript variants using 5’-RACE with cDNA derived from different tissues and analyzed all available TXNRD1-derived EST sequences. The results show that the core promoter governs transcription of the clear majority of TXNRD1 transcripts but also that alternative promoters may be activated under rare conditions or in specific cell types. Furthermore, extensive alternative splicing occured in the 5’-region of TXNRD1. In total twenty-one different transcripts were identified, potentially encoding five isoforms of TrxR1 carrying alternative N-terminal domains. One isoform encompassed a glutaredoxin domain, whereas another encoded a predicted mitochondrial localization signal. These results reveal that the human thioredoxin system is intriguingly complex. Cell specific transcription of the TXNRD1 gene encoding different isoforms of TrxR1 must hereby be taken into account in order to fully understand the functions of the human thioredoxin system.

Keywords

Thioredoxin reductase, alternative splicing, promoter, Sp1, Sp3

Introduction

The thioredoxin system is found in all organisms and consists of thioredoxin (Trx), thioredoxin reductase (TrxR) and NADPH. Trx is a small, well-characterized ubiquitous redox active protein, which is reduced by TrxR using NADPH as a source of reducing equivalents. Trx contains an active site with two redox active cysteine residues and participates in many different types of reactions including synthesis of deoxyribonucleotides, redox control of transcription factors and regulation of apoptosis (for reviews see [1,2]).

Mammalian thioredoxin reductases are selenocysteine-containing oxidoreductase flavoproteins with a broad substrate specificity, catalyzed by a carboxyterminal active site structure involving a conserved Gly-Cys-Sec-Gly tetrapeptide [3,4,5,6,7]. The enzymes reduce not only protein disulfides such as that in oxidized Trx, but also low molecular weight disulfide compounds, e.g. DTNB [8] and lipoic acid [9], as well as low molecular weight non-disulfide compounds such as selenite [10], alloxan [11], or peroxides [12]. Three different human genes that encode TrxR isoenzymes have been identified; the classical cytosolic TrxR1, a mitochondrial enzyme (TrxR2) [13,14,15] and one isoenzyme expressed mainly in testis [13]. These three proteins have the same overall domain structure. However, the testis-specific enzyme has an additional N-terminal glutaredoxin domain with a monothiol active site motif and can, in contrast to TrxR1 and TrxR2, also catalyze the direct reduction of both glutathione disulfide and thioredoxin. It has therefore been named "TGR" for Thioredoxin and Glutathione Reductase [16]. All mammalian TrxR transcripts contain a selenocysteine insertion sequence (SECIS) element in the 3'-untranslated region (UTR) necessary for selenocysteine incorporation [5,17,18,19,20]. In addition, the 3'-UTR of transcripts derived from TXNRD1 1) also contain AU-rich elements (AUREs) that lead to a rapid mRNA turnover [18,21]. AUREs are typically found in mRNA with rapid post-transcriptional up- and down-regulation such as cytokine, proto-oncogene, transcription factor, and other transiently expressed mRNAs [22]. The fact that TXNRD1 transcripts have functional AUREs is interesting since this enzyme is not transiently expressed only in certain cell types or under specific growth conditions, but is instead widely expressed in many diverse tissues [14,23,24] and cells [25]. We recently reported the initial characterization of the proximal promoter for human TXNRD1[26]. Deletion constructs were made and revealed that the promoter activity was maintained within the -115 to +167 region in all cell lines tested. This region was therefore considered as the core promoter region. The core promoter lacked classical TATA or CCAAT boxes but had an increased GC content. It contained a POU motif which was shown to bind the Oct-1 transcription factor and two GC rich regions that bound Sp1 and Sp3. Lack of classical TATA or CCAAT boxes, an increased GC content with functional Sp1 site(s) in the proximal promoter region, and a predicted CpG island close to the transcriptional initiation site are features typical of housekeeping genes [27,28]. Also Oct-1 is ubiquitously expressed and believed to govern the transcription of many housekeeping genes [29]. A housekeeping-type promoter in combination with AUREs in the 3’-UTR of TXNRD1 transcripts suggest an intricate regulation pattern for this enzyme [26]. A second level of regulation may furthermore derive from alternative splicing events.

We and others have shown that mammalian TrxR1 proteins may exhibit alternative splicing around the first exon [30,31,32]. At least three forms of transcripts differing in the 5’-end have been identified in mouse and rat and five forms in human [30,31,32]. One of the alternative murine splice variants contains an additional upstream in-frame ATG which could encode an N-terminally elongated protein of 67 kDa instead of the common 55 kDa form [31]. However, that mRNA form could not be found in humans [32] although a human TrxR1 variant protein with an apparent mass of 67 kDa was detected in the JPX9 cell line of T-cell origin [31].

In this study we have continued to analyse the human TXNRD1 core promoterwith point mutations in the previously studied Sp1/Sp3 and Oct1 binding sites and we identify one of the two Sp1/Sp3 sites as being important for the basal promoter activity. We have also cloned and sequence determined 5'-RACE variants of human TXNRD1 transcripts originating from different tissues and analyzed TXNRD1-specific EST sequences. We show that the clear majority of all TXNRD1 transcripts initiate at the site of the core promoter. However, we also found an extensive alternative splicing pattern in the 5’-region of TXNRD1. In total, we have here identified twenty-one different exon combinations and evidence for rare alternative promoters. Moreover, we identify five different N-terminal domains that may be encoded by these mRNA variants, three of which have not been described earlier. These results reveal that in spite of a housekeeping-type core promoter, transcription of TXNRD1 seems to be regulated in an exceedingly complex manner.

Materials and Methods

5'-RACEs of TrxR1 from human cDNA libraries

Marathon-Ready cDNA libraries (Clontech) from four different tissues (mammary gland, ovary, testis and thymus) were used to amplify the 5'-ends of human TrxR1. The libraries were used together with three gene specific primers (P1, 5'-GCT TCT ATC ATT CTG TCC CAA TCA TG-3'; P2, 5'-GAC ATG CTG AAG CTT TGT GTG ACC-3'; P3, 5'-GCA ACC CAC ATT CAC ACA TGT TCC-3') complementary to different parts of the TrxR1 cDNA as described in the text. The obtained products were analyzed on a 1% agarose gel, isolated and cloned into pGEM-T vectors (Promega). The plasmids were propagated in XL1-Blue Supercompetent cells (Stratagene) and sequenced. The sequences were deposited in Genbank with accession numbers given in Table I.

Sequence analyses

The GenBank EST database was searched for TXNRD1 specific sequences using the here identified 5’-RACE products as queries, which were aligned using the Pairwise BLAST Tool (NCBI) and/or the GeneJockeyII Sequence Processor (BIOSOFT). Protein characteristics were analyzed with PEPSTATS ( and prediction of cellular localization was analyzed with iPSORT and/or PSORT II ( Intron-Exon boundaries were inspected manually using the “show sequence” feature of the NCBI Map Viewer display of the human genome (

Construction of Sp1 and Oct1 mutants of the TXNRD1 core promoter

To construct mutants of the Sp1/Sp3 and the Oct1 sites in the core promoter of TXNRD1 the following primers were used: Sp1A, 5'-GCC CGC TCG GCG CAG TTT GTG GCT TCT CGT-3'; Sp1B, 5'-GCC CGC TCG GTTTAG GGC GTG GCT TCT CGT-3'; Sp1C, 5'-GCC CGC TCG GCTTAG TTC GTG GCT TCT CGT-3'; Oct1, 5'-AGC TTA CTA GGC AGCGAG CAT AGG TTG CC-3' with the underlined nucleotides introducing the desired point mutations. The QuickChange Site-Directed Mutagenesis kit (Stratagene) was used and a polymerase chain reaction (PCR) was run with the core promoter-luciferase construct called HA [26]as template using 16 cycles of 95oC 40 sec, 55oC 1 min, 68oC 12 min, according to the manufacturers protocol. The PCR products were treated with DpnI, re-ligated and propagated in XL1-Blue Supercompetent cells (Stratagene) and the sequences and desired mutations verified.

Cell cultures

The A549 (human lung carcinoma) and HeLa (human cervix carcinoma) cell lines were grown in Dulbecco's Modified Eagle's Medium supplemented with 10% heat-inactivated fetal calf serum, 2mM L-glutamine, 100 units/ml penicillin and 100 µg/ml streptomycin, in an atmosphere of 5% CO2 and 95% air at 37oC.

Determination of reporter gene activity

Approximately 3 x 104 cells per well in 24-well dishes were seeded 24 h before transfection using the mutated TXNRD1 promoter reporter constructs (0.5 µg) and co-transfection with 0.5 ng pRL-CMV plasmid. The transient transfections were performed using Lipofectamine Plus Reagent (GIBCO-BRL), with 2 µg lipofectamine and 3 µl Plus reagent for 4 h, according to the manufacturer's protocol. After 24 h the cells were washed with phosphate-buffered saline (GIBCO-BRL) and lysed in 100 µl Passive Lysis Buffer (Promega) at room temperature for 15 min. The firefly and renilla luciferase activities in 10 µl of the cell lysates were determined using the Dual-Luciferase Reporter Assay system (Promega) according to the manufacturer's instructions using a Turner Designs TD-20/20 luminometer.

Nuclear extracts

Nuclear extracts were prepared from 1 x 107 A549 or HeLa cells, grown in large culture plates, which were scraped into PBS and centrifuged for 10 min at 1800xg. The cells were resuspended in 600 µl hypotonic buffer (10 mM HEPES, pH 7.9), 1.5 mM MgCl2, 10 mM KCl, 0.2 mM PMSF, 0.5 mM DTT) and centrifuged at 1800xg for 5 min at 4oC. The pellet was resuspended in 400 µl hypotonic buffer and incubated for 10 min on ice whereupon the cells were homogenized with 10 strokes using a Dounce homogenizer with a B-type pestle. The homogenate was centrifuged for 15 min at 3300xg at 4oC, the supernatant was removed, and the nuclei in the pellet resuspended in 180 µl low-salt buffer (20 mM HEPES, pH 7.9, 25% glycerol, 1.5 mM MgCl2, 10 mM KCl, 0.2 mM EDTA, 0.2 mM PMSF, 0.5 mM DTT). KCl was added to a final concentration of 0.4 M while stirring gently with a pipette tip. Nuclear proteins were extracted for 30 min on ice with continuous gentle mixing. Upon a centrifugation for 30 min at 25 000xg at 4oC the supernatant with extracted proteins was frozen immediately at -80oC in aliquots, which were thawed gently on ice at time for use in electrophoretic mobility shift assays.

Electrophoretic mobility shift assays

The EMSA probes used (obtained from Invitrogen) were the following (sense sequence shown):

Sp1-control: 5'-CCCGCTCGGC GCAGGGCGTG GCTTCTCGTA GCCATTAGGA-3'

Oct1-control: 5'-TCTCAGCTTA CTAGGCAATT AGCATAGGTT GCCAGGGCTG-3'

Sp1A-mut: 5'-CCCGCTCGGC GCAGTTTGTG GCTTCTCGTA GCCATTAGGA-3'

Sp1B-mut: 5'-CCCGCTCGGT TTAGGGCGTG GCTTCTCGTA GCCATTAGGA-3'

Sp1C-mut: 5'-CCCGCTCGGC TTAGTTCGTG GCTTCTCGTA GCCATTAGGA-3'

Oct1-mut: 5'-TCTCAGCTTA CTAGGCAGCG AGCATAGGTT GCCAGGGCTG-3'

Double-stranded oligonucleotides were generated by annealing equi-molar complementary oligonucleotides in 50 mM Tris-HCl (pH 7.4), 1 mM EDTA, 100 mM NaCl and 13 mM MgCl2 with the following temperature program; 88oC 2 min, 65oC 10 min, 37oC 10 min, 25oC 5 min. The double-stranded oligonucleotides were end-labeled with [-32P]ATP (3000 mCi/mmol, DuPont NEN) using T4 polynucleotide kinase and the labeled probes were purified by Chroma-spin 30 (Clontech). For binding assays a mixture was prepared containing [32P]-labeled oligonucleotide (0.3 ng), 3 µg nuclear proteins, 1 µg poly(dIdC) (Amersham Pharmacia), adjusted to 20 µl with binding buffer (20 mM HEPES pH 7.9, 1.5 mM MgCl2, 0.2 mM EDTA, 1 mM DTT, 0.1 mM PMSF, 7.5% glycerol). Binding reactions were allowed for 15 min at 28oC whereupon a 15 µl aliquot of each reaction was loaded onto a 4% non-denaturing polyacrylamide gel and run in 1x Tris-Glycine buffer at 120 V. Following electrophoresis, gels were dried and autoradiographed.

Results

Transcriptional activity of the human TXNRD1 core promoter with point mutated Sp1/Sp3 and Oct-1 sites

In a previous study we cloned and characterized the promoter region of human TXNRD1[26] and found a predicted POU domain (-98 to -90) in the core promoter which was functional in binding the Oct-1 transcription factor. Immediately upstream of the transcription start point a GC-rich region was present (-17 to -7), which also bound the Sp1 and Sp3 transcription factors [26]. Here we wished to verify the function of these transcription factor motifs in human cells. For this, the Oct-1 binding site (-98 to -90) was mutated from ATTAGCATA to GCGAGCATA (Oct-1m) and in case of the Sp1/Sp3 tandem-binding site (-17 to -7) three different mutants were made, with the sequence changed from GCGCAGGGCGT to GCGCAGTTTGT (Sp1Am), to GTTTAGGGCGT (Sp1Bm), or to a combination of the two mutated consensus motifs GCTTAGTTCGT (Sp1Cm). The mutated oligonucleotides were first analyzed using electrophoretic mobility shift assays with nuclear extracts from either A549 or HeLa cells. The electrophoretic mobility shift caused by Oct-1 was abolished using the Oct-1mut oligonucleotide, whereas the shifts caused by Sp1/Sp3 could still be detected with the Sp1Bmut variant but not with Sp1Amut or Sp1Cmut (Fig 1). Reporter constructs for the core promoter having the corresponding mutations were then used for transfections of A549 and HeLa cells and the transcriptional activity of the luciferase reporter gene was measured. As a negative control a construct with the upstream region cloned in reverse orientation (HB) was used, known to be transcriptionally inactive [26], whereas the wildtype reporter construct, called HA, served as positive control. In both cell lines, the Sp1Bmut and Oct-1mut constructs showed a luciferase activity comparable with the wildtype HA construct, whereas Sp1Amut and Sp1Cmut gave reduced transcriptional activities. The luciferase activities of Sp1Amut and Sp1Cmut in A549 cells were more decreased than in HeLa cells, with a 50% decrease in A549 cells compared to a 25% decrease in HeLa cells (Fig 2). These results show that the POU site and the second Sp1/Sp3 consensus motif covered by the Oct-1mut and Sp1Bmut constructs are dispensable for basal TXNRD1 transcription in A549 and HeLa cells. The Sp1/Sp3 site covered by the Sp1Amut on the other hand contributes to a major part of the basal TXNRD1 transcriptional activity in these cells.

Cloning of cDNA variants of TXNRD1-derived transcripts differing in the 5'-region

As mentioned in the Introduction, we and others have reported the existence of up to five separate splice variants of the 5’-UTR of human TXNRD1. Those variants were mostly discovered only on the basis of database searches with identification of EST sequences derived from high-throughput screening projects. Since different 5’-UTR sequences in theory may reveal the existence of alternative promoters or transcriptional start sites, we decided to focus on the possible presence of 5’-UTR variants using an experimental approach. The results, as described next, confirmed that the hitherto described core promoter is the major promoter for human TXNRD1, but also revealed a surprisingly high number of alternative 5’- variants.

In order to examine human TXNRD1 5'-region variants, we screened four separate cDNA libraries from different tissues (mammary gland, testis, ovary and thymus). Three primers complementary to three strategic parts of the TXNRD1 sequence were designed for these 5' RACE's. One primer (P1) was based on the region downstream of the well known ATG (present in what we subsequently named Exon III), another primer (P2) was based on an exon covering an alternative upstream in-frame ATG (Exon I) resulting in the previously published “KDRF” variant of TrxR1 [21], whereas the third primer (P3) was complementary to the nucleic acid sequence encoding the N-terminal disulfide/dithiol active site CVNVGC (Exon V). These three primers generated several separate TXNRD1-specific RACE products with different lengths and pattern in the examined tissues. In total, we characterized fourty-four different PCR products, the analysis of which revealed eighteen separate types of transcripts for TXNRD1. We then complemented this information with an extensive GenBank database search, which identified EST clones that supported these results. Furthermore, some EST sequences indicated the presence of additional splice variants not identified in our experimental approach, which together gave a total of twenty-one separate splice variants. All exons could be identified in the human genomic sequence covering the TXNRD1 gene on chromosome 12. As an aid to discuss and describe the hereby identified transcript variants, we will use the following nomenclature:

  • Exons present in the most common transcripts are given roman numerals (I, II, III, IV, V, …) in order of appearance in the genome. The transcriptional start site is indicated in parenthesis after the first exon, with “1” either being the previously denoted transcriptional start site for exon I [26] or the first nucleotide after a splice site of a regular exon, in case a transcript initiates with another exon than exon I. Variant exons extended at the 5’-end are denoted with a “v” prefix (e.g. vII), whereas exons with alternative donor splice sites are denoted with a small letter suffix (e.g. the exon Ia is a shorter version of Ib due to the use of an alternative donor splice site).
  • Rare exons are named with Greek letters and consecutive subscript roman number (e.g. I, II, III, … or -I, -II, -III, …), where the number gives the order of the exons as counted from the core promoter; a negative number indicates an upstream location.
  • Transcripts containing only the most common exons are called  transcripts, those containing  exons are named  transcripts, whereas  transcripts contain  exons. Transcripts that differ due to alternative splicing patterns are given consecutive numbers (e.g. 1, 2, 3, …)
  • Potential protein isoforms encoded by the different splice variants having alternative N-terminal domains are specifically named as TXNRD1_v1, TXNRD1_v2, etc., following the guidelines of the HUGO Gene Nomenclature Committee [33].

All the TXNRD1-derived 5’ variant transcripts hence identified are listed in Table I and schematically drawn in Figure 3. The intron-exon boundaries are given in Table II and variant N-terminal domains are shown in Figure 4.