HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 25, 22460-22468, June 21, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/25/22460    most recent M202286200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ma, X. Articles by Kalvakolanu, D. V. Search for Related Content PubMed PubMed Citation Articles by Ma, X. Articles by Kalvakolanu, D. V. Social Bookmarking                      What's this?

Mutational Analysis of Human Thioredoxin Reductase 1

EFFECTS ON p53-MEDIATED GENE EXPRESSION AND INTERFERON AND RETINOIC ACID-INDUCED CELL DEATH^*

Xinrong Ma, Junbo Hu^§, Daniel J. Lindner^¶, and Dhananjaya V. Kalvakolanu

From the Greenebaum Cancer Center, Department of Microbiology & Immunology, Molecular and Cellular Biology Program, University of Maryland School of Medicine, Baltimore, Maryland 21201 and the ^¶ Taussig Cancer Center, Cleveland Clinic Foundation, Cleveland, OH 44195

Received for publication, March 8, 2002, and in revised form, April 10, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The interferon (IFN)- and all-trans-retinoic acid combination suppresses tumor growth by inducing apoptosis in several tumor cell lines. A genetic technique permitted the isolation of human thioredoxin reductase (TR) as a critical regulator of IFN/all-trans-retinoic acid-induced cell death. Our recent studies have shown that TR1:thioredoxin 1-regulated cell death is effected in part through the activation of p53-dependent responses. To understand its death regulatory function, we have performed a mutational analysis of TR. Human TR1 has three major structural domains, the FAD binding domain, the NADPH binding domain, and an interface domain (ID). Here, we show that the deletion of the C-terminal interface domain results in a constitutive activation of TR-dependent death responses and promotes p53-dependent gene expression. TR mutant without the ID still retains its dependence on thioredoxin for promoting these responses. Thus, our data suggest that TR-ID acts as a regulatory domain.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Interferons (IFNs)¹ exert antitumor effects by inducing the expression of a number of cellular genes using the Janus tyrosine kinase-signal transducers and activators of transcription (STAT) pathways (1, 2). A higher susceptibility of IFN- receptor^/ and STAT1^/ mice to chemical carcinogenesis as compared with their wild-type counterparts and a failure of syngeneic mice to reject the IFN- receptor^/ and STAT1^/ tumors underscore the importance of IFNs in tumor growth control (3). Similarly, two IFN-regulated transcription factors, IRF-1 and IRF-8 (IFN consensus sequence binding protein), act as tumor growth suppressors (4, 5) because mutations in these genes cause leukemias (6, 7). In rodent cells, IFN-stimulated transcription factors of the p200 family control cell cycle progression (8, 9). IFNs also down-regulate c-myc expression, activate tumor suppressor pRb, and inhibit transcription factor E2F to inhibit cell cycle progression in human cell lines (10-12). Although a great deal is known about IFN signaling pathways and the transcription factors involved, very little is known about the gene products that mediate the tumor-suppressive pathways employed by IFNs. Additionally, despite their beneficial therapeutic effects in certain leukemias, IFNs are marginally active in the therapy of solid tumors (13, 14). Clinical and experimental models have shown that the combination of IFNs with retinoids, a class of vitamin A derivatives, yields a highly effective growth-suppressive effect in several solid tumors (15-17). All-trans-retinoic acid (RA), a vitamin metabolite, inhibits the growth of promyelocytic leukemias, and teratocarcinomas in vitro (18). Two structurally similar but genetically distinct classes of transcription factors, the retinoic acid receptor and the retinoid X receptor, mediate retinoid-induced growth suppression (19). One such receptor, retinoic acid receptor , appears to be a tumor suppressor (20, 21). However, the identity of retinoid-regulated growth-inhibitory gene products is unknown.

Our earlier studies showed that the combination of IFN- and RA, but not the single agents, causes cell death in vitro and suppresses tumor growth in vivo (17). Using a genetic technique, we have recently identified several genes associated with retinoid-IFN-induced mortality (22). Human thioredoxin reductase (TR) 1, a redox regulatory enzyme (23, 24), was identified as one of the genes associated with retinoid-IFN-induced mortality (22). Subsequent studies have shown that TR and its substrate, Trx, activate cell death by modulating the activity of caspase-8 and tumor suppressor p53 (25-27). To further understand the structure-function relationship of TR to these processes, we have performed a mutational analysis. A comparative analysis of the primary structure of this enzyme with other redox enzymes led to the assignment of three major modules, the FAD, NBD, and ID (23, 24). Whereas the FAD and NBD are critical for redox function and are well conserved among the TRs from all sources, the ID is unique to mammalian TR. ID has been suggested to act as a dimerization surface to generate functional TR (23, 24). However, the functional significance of ID has not been fully appreciated. Here we show that removal of ID enhances death-stimulatory activity of TR and p53-dependent gene expression.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Reagents-- Restriction and DNA-modifying enzymes (PerkinElmer Life Sciences); G418 sulfate, isopropyl-1-thio--D-galactopyranoside, and LipofectAMINE PLUS (Invitrogen); nylon membranes, ECL reagents, and horseradish peroxidase coupled to anti-rabbit or anti-mouse antibodies (Amersham Biosciences); human IFN-_ser (Berlex Inc.); and mouse monoclonal antibodies against actin, FLAG epitope (Sigma), p53 (Oncogene Science Inc.), and myc-epitope (Zymed Laboratories Inc.) were employed in these studies. Rabbit polyclonal antibody against the C-terminal peptide of TR1 was described previously (22). Fresh stocks of all-trans-retinoic acid (Sigma) were prepared in ethanol and added to cultures under subdued light.

Cell Culture-- MCF-7 cells were cultured in phenol red-free Eagle's minimal essential medium supplemented with 5% charcoal-stripped fetal bovine serum and 10¹¹ M estradiol during treatment with IFN- and RA. MCF-7 cells stably transfected with wild-type and mutant forms of Trx have been described previously (25). The mutant Trx bears serine residues in the place of cysteines at positions 32 and 35 (28). MCF-7 cells stably transfected with mammalian expression vector pCMV-neo (MCF-7 neo) or the same vector with the E6 gene of human papilloma virus type-16 (MCF-7 E6) were provided by A. J. Fornace Jr. (National Cancer Institute, Bethesda, MD) (29). The loss of p53 function in MCF-7 E6 cells has been demonstrated previously (29, 30). These cells were grown in phenol red-free media 24 h before treatments were initiated. DLD human carcinoma cells, which lack endogenous p53, were a gift from Bert Vogelstein (Johns Hopkins University Oncology Center, Baltimore, MD).

Plasmids-- Mammalian expression vector pCMV-FLAG bears a FLAG epitope sequence in its multiple cloning region. An in-frame insertion of any cDNA lacking the N-terminal methionine into this vector generates the protein with a FLAG epitope tag at the N terminus. Mammalian expression vector pCXN2-myc carries a C-terminal myc epitope tag. Cloning of an insert without a "stop" codon between the 5' EcoRI and 3' KpnI sites of this vector permits the addition of a myc tag to the expressed protein. The p53-Luc reporter carries eight copies of the p53 binding element cloned upstream of the SV40 early promoter in the pGL3 basic vector (Promega) was described earlier. Wild-type and mutant p53 (R175H) cloned in the pCMV expression vector have been described elsewhere (31, 32). A luciferase reporter driven by human Bax promoter, Bax-Luc, was provided by Carol Prives (Columbia University, New York, NY) (33).

Generation of TR Mutants-- Gene-specific primers bearing specific restriction enzyme-cutting sites (for facilitating subcloning) and AmpliTaq gold enzyme (Roche Molecular Biochemicals) were employed in PCR for generating TR mutants. All primers used in this study are listed in Table I. Two separate sets of primers were used to generate myc- and FLAG epitope-tagged constructs. Construction of a myc-tagged full-length TR was described in our earlier publication (26). At first, the myc-tagged mutants were generated, which served as templates for generating FLAG-tagged mutants. Twelve cycles of PCR were performed to avoid the emergence of unwanted mutants due to polymerase errors. Mutants were sequenced to verify their identity.

Mutant -FAD was amplified using myc primers 2 and 3 with wild-type TR cDNA as template. -ID was generated using myc primers 1 and 4. To generate the -NBD mutant, four primers were used. The myc primer 1 and -NBD primer 1 were used to amplify the FAD domain. The myc primer 2 and -NBD primer 2 were used to amplify the ID region. Oligonucleotides used for amplification of the FAD and ID regions have a BamHI site at their 3' and 5' ends, respectively. The myc primers 1 and 2 bore EcoRI and KpnI sites. The FAD product was digested with EcoRI and BamHI, and the ID product was digested with BamHI and KpnI and then purified. The products were combined with pCXN2-myc vector predigested with EcoRI and KpnI in a three-way ligation reaction. The final -NBD construct has two non-template-derived amino acids, a glycine and a serine, at the junction of FAD and ID, due to a BamHI site present in the amplifying primers. Constructs f-TR and f--NBD were generated using FLAG primers 1 and 2, with the corresponding myc-tagged constructs as templates. The f--ID and f--FAD mutants were generated using FLAG primers 1 and 4 and FLAG primers 3 and 2, respectively. f-tagged truncated mutants of -ID were generated using the indicated reverse primers and FLAG primer 1, with wild-type TR as template. Point mutants were generated using -ID80 as a template. For example, to generate the T193A mutant, a reverse primer and a forward primer bearing the same mutation were used in a three-step PCR. FLAG primer 1 and reverse primer (mutagenic) were used in the first PCR reaction. Forward primer (mutagenic) and -ID80 R primer were used for the second PCR. Purified PCR products from the first and second PCR were mixed, denatured, and annealed. This mixture now served as template for FLAG primer 1 and -ID80 R primer to generate the final product. The final PCR product was digested with EcoRI and KpnI and ligated to pCMV-FLAG. The other mutants were generated in a very similar manner, using appropriate mutant primers.

Cell Growth Assay-- Cells (2000 cells/well) were seeded into 96-well plates. Drugs were added, and growth was monitored using a colorimetric assay (34). Each group of treatments had eight replicates. Cells were fixed with 10% trichloroacetic acid at the end of the experiment and stained with 0.4% sulforhodamine B (Sigma). The bound dye was eluted with 100 µl of Tris-HCl (pH 10.5), and absorbance was monitored at 570 nm. One plate was fixed with trichloroacetic acid, 10 h after plating. Absorbance obtained with this plate was considered as 0% growth. Absorbance obtained with untreated cells was considered as 100% growth. An increase and decrease of A₅₇₀ values in the experimental wells relative to the 0% value indicate cell growth and death, respectively.

Death Assays-- Cell death was determined using annexin-V binding assays. After treatment with IFN/RA, cells were stained using a commercially available kit (Trevigen Inc.) per the manufacturer's recommendation. FITC-positive cells were considered apoptotic and quantified using flow cytometry.

Gene Expression Analyses-- Transfection, -galactosidase and luciferase assays, SDS-PAGE, and electrophoretic mobility shift analyses (EMSAs) were performed as described in our previous publications (25-27). The total amount of transfected DNA (1.0 µg) was kept constant by adding corresponding empty expression vector DNA, where required. In general, 0.2 µg of luciferase and 0.2 µg of TR mutant were co-transfected. CMV -galactosidase reporter (0.1 µg) was used as an internal control for normalizing variations in transfection efficiency. Electrophoretic mobility shift assay with p53 oligonucleotides was performed as described previously (25-27).

Western Blot Analysis-- Equal quantities of cell extracts were separated on 12% SDS-PAGE and Western blotted onto nylon membranes. Specific first antibodies were incubated with the blots as described in our previous publication (22). These blots were washed and incubated with an appropriate second antibody tagged with horseradish peroxidase. Protein bands were visualized using a commercially available enhanced chemiluminescence kit (ECL; Amersham Biosciences).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Generation of TR Mutants-- We have shown previously that overexpression of a catalytically inactive TR1 or a redox inactive Trx1 in human tumor cell lines imparts resistance to IFN/RA-induced cell death (25, 26). In contrast, a wild-type TR1 and Trx1 promoted cell death under the same conditions. However, the role of other TR domains in cell growth control is unknown. To understand the relationship between structural domains of TR and cell death regulation, we have generated new TR mutants. Using PCR we generated three mutants, -FAD, -NBD, and -ID, which lacked the FAD binding domain, the NADPH binding domain, and the interface domain, respectively. Because no domain-specific antibodies are available for TR1, we have cloned the PCR products into mammalian expression vectors, pCMV-FLAG or pCXN2-myc. Proteins expressed from pCMV-FLAG and pCXN2-myc will bear an N-terminal FLAG- and a C-terminal myc epitope tag, respectively. Wild-type TR produces a polypeptide with a theoretical molecular mass of 54.7 kDa. However, it migrates as a ~58-kDa protein on SDS-PAGE due to posttranslational modifications. The -FAD, -NBD, and -ID constructs are expected to yield 40.1-, 33.2-, and 27.6-kDa peptides, respectively. The mutants were transiently transfected into human breast carcinoma cell line MCF-7 to check for the production of proteins of proper size. Cell lysates were prepared and Western blotted using either FLAG- or myc epitope-specific monoclonal antibodies. Indeed, all mutants can be expressed to a comparable level upon transfection (Fig. 1, B and C). Both tags were used only to demonstrate that either N-terminal or C-terminal tags have no effect on protein function. Furthermore, pCXN2-myc has a G418 resistance marker (G418^r) for selecting stably transfected cells, which is absent from pCMV-FLAG. The FLAG tag is shorter than the myc tag by about 8 amino acids.

View larger version (53K): [in this window] [in a new window]   Fig. 1.   Expression of TR mutants. Mutants were generated using PCR as described under "Materials and Methods." A, diagrammatic representation of the structures of TR mutants. B and C show the Western blot (WB) analyses. MCF-7cells were transiently transfected with the indicated plasmids (1 µg), and equal amounts of lysates were separated on a 10% SDS-PAGE and Western blotted. The blots were probed with indicated antibodies.

Removal of ID Enhances the Cell Death Effects of TR-- To test the effect of TR mutants on cell growth, we first generated cell lines that stably express them. Because the pCXN2-myc vector also carries a G418^r marker that permits the selection of stable transfectants, we have used the myc-tagged constructs for this purpose. As observed earlier, transfection of a wild-type TR resulted in the formation of fewer colonies than the vector. The -FAD construct gave rise to marginally more colonies than the control vector-transfected cultures, indicating its inhibitory effect on cell death. The -NBD mutant and vector produced a comparable number of colonies. Interestingly, the -ID mutant yielded 75% and 50% fewer G418^r colonies than the vector- and TR-transfected cultures, respectively (Fig. 2A). The G418^r colonies in each group were pooled and used in the experiments described below to avoid a clonal bias. In the next experiment, the effect of TR mutants on cell growth was determined using a colorimetric assay (34), where cell growth was quantified on the basis of the amount of sulforhodamine B dye bound to cells (Fig. 2B). This method correlates well with Coulter counting and determination of cell number. Whereas the -FAD-expressing cells grew slightly faster than the vector-transfected cultures, the TR and -ID transfectants grew relatively slower. Expression of -ID caused significantly slower growth compared with TR. Growth of -NBD transfectants was comparable to that of vector transfectants. These observations suggest that removal of ID converts TR into a significantly more potent inhibitor of cell growth. Similar results were obtained with f-tagged TR mutants (data not shown).

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Effects of TR mutants on cell growth. A, effect of TR mutants on G418r colony formation. MCF-7 cells were transfected with equimolar amounts of pCXN2-myc vector expressing various mutants After 3 weeks of selection with G418 (1 mg/ml) in growth medium, surviving colonies were counted. Each bar represents the mean ± S.E. of triplicates. B, effects of TR mutants on cell growth. An equal number of cells (2000 cells/well) stably transfected with various TR mutants were plated, and cell growth was monitored after 5 days using the sulforhodamine B binding assay (34). Each bar represents the mean ± S.E. of eight replicates. Cell growth was monitored and quantified by measuring the absorbance of bound dye at 575 nm. C, expression of TR mutants after stable transfection in MCF-7 cells. Total protein (65 µg) from the indicated cell lines was Western blotted and probed with anti-myc antibody. This blot was stripped and probed (bottom panel) with actin-specific antibodies. D, cells treated with IFN/RA were stained with annexin-V as described under "Materials and Methods." The percentage of annexin-V-positive cells was quantified using fluorescence-activated cell-sorting analysis. Each bar represents the mean ± S.E. of triplicates.

To demonstrate that these differential effects on cell growth were due to the expression of mutants in the stable transfectants, cell lysates were examined for the expression of transgenes by Western blot analysis with myc tag-specific antibodies (Fig. 2C). Although all mutants were expressed in the transfectants, wild-type TR and -ID were expressed to a lesser extent. In fact, the expression of -ID was lost with further passages of the transfectants (data not shown), indicating its strong anticellular effects. To demonstrate a functional relationship between cell death and mutant expression, the stable transfectants were exposed to the IFN/RA combination (IFN/RA) and then stained with FITC-labeled annexin-V, a maker for apoptotic death (35). A higher FITC-positive staining indicates higher apoptosis. FITC-positive cells were quantified using flow cytometry (Fig. 2D). The TR and ID transfectants exhibited a significantly higher sensitivity to IFN/RA-induced cell death compared with the other mutants. The -ID-expressing cells became FITC-positive 2-2.5-fold higher than TR-expressing ones. The FAD mutant acted as an inhibitor of apoptosis because cells expressing it were less FITC-positive than the vector-expressing ones. Together, these data indicate that ID of TR attenuates its proapoptotic effects.

Expression of -ID Has No Effect on Endogenous TR and p53-- To rule out the possibility that the expression of TR or -ID somehow altered the endogenous TR levels to mediate these differential effects, we next determined the levels of endogenous TR protein by Western blotting with antibodies specific for TR. Because only -ID exhibited hyper-death-stimulatory effects, we have selected it for additional studies and compared its effects to full-length TR. As shown in Fig. 3A, neither the -ID nor wild-type TR had an effect on endogenous TR because its level was comparable between the vector- and mutant-transfected cells. In wild-type TR-transfected cells, a slow migrating band above the TR band was detected. It corresponds to the TR protein derived from the transgene and migrates slower because of the presence of an epitope tag. This band is absent in the vector- or -ID-expressing cells. Because the TR antibody was directed against a peptide in the C terminus, -ID could not be detected.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Equal amounts of total cell lysate (85 µg) from the indicated cell lines were immunoblotted and analyzed with specific antibodies. A, expression of endogenous TR protein. Arrowhead indicates the position of transgene-derived TR. This antibody does not recognize protein derived from -ID. This blot was stripped and probed with anti-actin antibodies (bottom panel of B). B, expression of endogenous p53 protein. WB, Western blot.

We have shown previously that the cell death effects of TR were due in part to its ability to modulate tumor suppressor p53-dependent responses (25-27). Therefore, we examined the possibility that a rise in endogenous p53 levels of mutant transfectants relative to vector-expressing cells occurred. Expression of -ID or wild-type TR did not significantly affect p53 levels as revealed by a Western blot analysis with an antibody that specifically detects a wild-type p53 protein (Fig. 3B).

Effect of the IFN/RA Combination on p53-regulated Gene Expression-- We have reported previously that TR modulates p53-dependent cell death via an up-regulation of gene expression (27). To test the influence of TR mutants on p53-stimulated gene expression, we have performed the following experiments. First, we wanted to know whether -ID induces the expression a reporter gene driven by p53 response element. MCF-7 cells were transfected with a p53RE-Luc reporter. Along with the reporter pCMV-FLAG, wild-type f-TR or f--ID mutant was co-transfected (Fig. 4A). Because the FLAG tag is shorter and yielded a slightly better expression in transient assays, we used the FLAG-tagged mutants for the following studies. Nevertheless, the myc-tagged mutants exhibited properties similar to those of the FLAG-tagged ones (data not shown). After transfection, cells were treated with IFN/RA, and luciferase activity was measured. IFN/RA induced luciferase expression in the vector transfectants. In TR transfectants, basal luciferase activity was elevated, and it was further strongly induced by IFN/RA. -ID co-expression strongly enhanced luciferase expression, and it was only slightly but significantly stimulated further by IFN/RA. -ID induced slightly higher luciferase expression than the IFN/RA-treated, TR co-expressed control. Previously, we have shown that Bax, a p53-responsive mRNA, and its protein are induced in the presence of wild-type TR and inhibited in the presence of a catalytically inactive mutant. Because p53-Luc used in this experiment contained a synthetic promoter, we next explored whether -ID exerted a similar effect on a native promoter. A luciferase reporter driven by human Bax has been shown to respond to p53 (33, 36). Therefore, we have employed it in the next experiment. Fig. 4B shows the data obtained. -ID constitutively activated this promoter in a manner similar to p53-Luc. TR, on the other hand, required treatment with IFN/RA to exert a similar effect. Thus, a synthetic and the native promoter respond similarly to -ID.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Effect of IFN/RA on p53-dependent gene expression. A, MCF-7 cells were treated with IFN/RA after transfection with p53-Luc, CMV--galactosidase, and TR mutants. Cell extracts were measured for luciferase and -galactosidase activity. Luciferase activity was normalized to the -galactosidase activity. Each bar represents the mean ± S.E. of triplicates. + indicates treatment with IFN- (500 units/ml) and RA (1 µM) for 16 h. B, effect of IFN/RA on the Bax promoter. MCF-7 cells were transfected with Bax-Luc and CMV--galactosidase plasmids. + indicates cells that were stimulated with IFN/RA. C, effect of TR mutants on p53 binding to DNA. Cell extracts from IFN/RA-stimulated cells (24 h) were incubated with a 32P-labeled oligonucleotide bearing the consensus p53 binding site.  and + indicate no treatment and IFN/RA treatment, respectively. 50X cold indicates that the -ID cell extract was incubated with an excess (50×) unlabeled oligonucleotide before use in EMSA. Thirty µg of nuclear extract was used in this experiment.

Previously, we have shown that in the presence of a wild-type TR, IFN/RA treatment enhances the DNA binding of p53, and a catalytically inactive TR blocks it. Therefore, we examined the DNA binding of p53 in cells stably expressing vector, wild-type TR, and -ID. Our previous studies have shown that exposure of cells to IFN/RA for 24-28 h, a time when significant apoptosis can be detected, is optimal for detecting p53 binding by EMSA without causing a rise in p53 levels (25-27). Whereas an IFN/RA-dependent induction of p53 binding occurred in the TR-expressing cells, a constitutive binding of p53 to the response element was observed in cells expressing -ID (Fig. 4C). It was slightly enhanced by IFN/RA. This observation is consistent with the luciferase expression data. This band was competed out by a cold p53 binding element. We have shown earlier that this band is not competed out by a mutant p53 oligonucleotide and that it is supershifted by a polyclonal antibody against p53 (25-27).

p53 Is Necessary for a Hyperstimulating Effect of -ID on Gene Expression-- A critical role for p53 in TR-regulated cell death effects was examined using DLD human colon carcinoma cells, which lacked the endogenous p53 (37). Reintroduction of p53 causes the death of these cells (37). Introduction of a p53-Luc reporter along with an empty vector did not cause an up-regulation of luciferase activity. -ID alone had no effect on luciferase expression. In the wild-type p53-transfected cells, luciferase expression was stimulated 3.5-4.0-fold over the vector-transfected cells. Luciferase expression was induced further significantly in the presence of -ID. Furthermore, the same reporter was not induced in cells transfected with a mutant p53, and co-transfection of -ID had no effect (Fig. 5A). These data show that p53 is obligatory for an inductive effect of -ID on the luciferase reporter.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   p53 is required for the gene-stimulatory effect of -ID. In A, the human DLD colon carcinoma cell line (p53-null) was transfected with the indicated plasmids along with p53-Luc and CMV--galactosidase reporters. + indicates the presence of the indicated plasmid in the transfection mixture. In B, MCF-7 neo and MCF-7 E6 cells were transfected with p53RE-luciferase in the presence of various indicated plasmids. The total amount of DNA transfected into the cells was kept constant (1.0 µg) by adding the pCMV-FLAG vector, where required.

Human breast carcinoma cell line MCF-7 carries a wild-type p53 allele (38). The endogenous p53 can be inactivated by targeting it to ubiquitin-dependent proteolysis upon stable expression of the human papilloma virus E6 gene (29, 30). Such epigenetic down-regulation confers a p53-null property to the MCF-7 E6 cell line. The control cell line MCF-7 neo carries an empty expression vector. Both cell lines were transfected with a p53RE-Luc reporter along with empty vector, TR, or -ID construct. As shown in Fig. 5B, transfection of wild-type TR caused an elevation of luciferase gene expression in MCF-7 neo cells. -ID also induced luciferase expression, which was significantly higher than the wild-type TR. p53-dependent gene expression was induced by IFN/RA treatment in the vector-transfected cells. In the presence of wild-type TR, IFN/RA further induced luciferase activity strongly. IFN/RA caused only a marginal increase in luciferase expression in the -ID-transfected cells. Thus, deletion of ID permits a constitutive activation of p53-dependent gene induction. The same mutant, when introduced into MCF-7 E6 cells, failed to promote luciferase expression. These results show that p53 is critical for TR-mediated gene induction.

Trx Is Required for Hyperactivating p53-dependent Gene Expression-- Because mammalian TR exhibits a wide substrate range (23, 24), we next examined whether the hyperactivation of p53-dependent responses by -ID was a result of its shift from the use of its native substrate Trx1. Therefore, we next determined whether the -ID mutant activates p53-dependent gene expression in MCF-7 cells stably expressing a mutant Trx, which lacks the critical cysteines (at positions 32 and 35) for its redox function (25). For this purpose, we have employed three MCF-7 cell lines stably expressing the vector (V), wild-type Trx1 (W), or mutant Trx1 (M). Co-transfection of pCMV2-FLAG with p53-Luc had no effect on luciferase activity in the V, W, and M cell lines (Fig. 6). However, co-transfection of wild-type TR elevated the basal expression of p53-Luc in V cells, which was further stimulated in W cells. However, TR failed to augment p53-dependent gene expression in M cells. A similar pattern of gene regulation was obtained with -ID mutant in V, W, and M cells. The only major difference between the -ID and TR is that -ID enhanced the gene expression to a higher level than TR. Thus, Trx is required for the stimulatory effect of -ID on p53-inducible expression.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Trx1 is required for the hyperstimulation of p53-dependent gene expression. MCF-7 cells stably expressing control vector (V), wild-type (W), and redox inactive Trx1 (M) were transfected with TR mutants along with p53-Luc and CMV--galactosidase reporters. Luciferase activity was quantified as described in the Fig. 4 legend. + indicates the presence of that specific plasmid in the transfection mixture.

Minimal Region of TR Protein Required for Stimulating p53-dependent Gene Expression-- Based on the above results, we next determined the minimal region of -ID required for stimulating p53-dependent gene expression. Serial deletion mutants, each lacking a specific number of amino acids from the C terminus of -ID, were generated using PCR. These mutants (-ID80, -ID70, -ID60, -ID50, and -ID30, which lacked 34, 44, 54, 64, and 95 amino acids, respectively), were expressed as N-terminal FLAG-tagged proteins using pCMV-FLAG. These mutants yielded 23.6, 22.6, 21.6, 20.6, and 17.1 peptides, respectively. A Western blot analysis of transfected cell extracts with FLAG tag-specific antibodies showed the expression of these mutants (Fig. 7A). All mutants were expressed equivalently. We next determined the effect of these mutants on p53-dependent gene expression (Fig. 7B). The -ID80 mutant was better than -ID70 at augmenting the reporter gene expression constitutively, although -ID70 still retained a significant amount of stimulatory effect. The other mutants lost their stimulatory effects on p53-dependent gene expression. Although a slight stimulatory effect of IFN/RA was found on the -ID80 mutant, it was lost with -ID70. These data suggest that a domain between -ID80 and -ID60 regulates the constitutive stimulatory effect on p53-regulated gene expression.

View larger version (29K): [in this window] [in a new window]   Fig. 7.   Minimal region of -ID required for its hyperstimulatory effects on p53-responsive gene expression. A, expression of mutants in MCF-7 cells. Cells in 6-well dishes were transfected with 0.5 µg of indicated plasmids, and whole cell lysates were prepared. A comparable quantity of protein (50 µg) from each sample was employed for Western blot analysis using FLAG-specific antibodies. B, effect of the mutants shown in A on p53-dependnent luciferase reporter. The indicated plasmids were co-transfected with p53-Luc and CMV--galactosidase reporters, and luciferase expression was analyzed. + indicates treatment with IFN- (500 units/ml) and RA (1 µM) for 16 h.

Residues Critical for the Stimulatory Effect on p53-dependent Transcription-- In the next set of experiments, we determined the residues critical for the hyperstimulatory effect of -ID on p53-dependent gene expression. Based on the fact that -ID80 exhibits a strong constitutive effect on p53-dependent gene expression (Fig. 7B), we engineered new point mutants that lack specific residues. Because the -ID70 mutant has a significant stimulatory effect, we reasoned that residues might lie within it. Primary sequence analysis of this region revealed a NADPH binding domain. There are potential residues that can be phosphorylated in this region. These include a serine at 199, a threonine at 193, and two tyrosine residues at 187 and 200. Two of these are present within the NADPH binding motif. While mutating tyrosine 187, we simultaneously converted the adjacent cysteine residue at 189 into an alanine. We have thus generated three new mutants: 1) Y187F/C189A, 2) T193A, and 3) S199A/Y200F. These mutants were cloned into pCMV-FLAG, and their expression was verified by Western blot analysis of the transfected cell extracts (Fig. 8A). The mutants were co-transfected with a p53-luciferase reporter and analyzed for their stimulatory effect on the promoter. Mutants Y187F/C189A and T193A completely lost their stimulatory effect on p53-dependent gene expression. However, mutant S199A/Y200F retained its stimulatory effect, comparable to that of -ID80 (Fig. 8B). Lastly, single mutants of Y187F and C189A activated the p53-dependent gene expression like -ID, indicating that both amino acids play a crucial in the regulation of p53-dependent gene expression.

View larger version (47K): [in this window] [in a new window]   Fig. 8.   Residues required for the hyperstimulatory effect of -ID on p53-mediated gene expression. A shows the expression of the point mutants indicated above it. This experiment is similar to that in Fig. 7A. B shows the effect of point mutants on p53-dependent gene expression. Transfection and reporter gene analysis were similar to those in Fig. 7B. C, EMSA for p53 activation. Stable cell lines expressing FLAG-tagged TR mutants (indicated above the lanes) were treated with IFN/RA for 30 h, and EMSA was performed. + indicates treatment with IFN- (500 units/ml) and RA (1 µM). The location of the p53 band is indicated.

The influence of point mutants on p53 activation was analyzed by EMSA. Stable cells lines expressing FLAG-tagged mutants were generated (see Fig. 9). Four cell lines that expressed -ID, Y187F/C189A, T193A, or S199A/Y200F were utilized for EMSA (Fig. 8C). Whereas -ID and the S199A/Y200F mutant caused an elevation of p53 binding to the response element, the Y187F/C189A and T193A mutants did not. IFN/RA did not cause an activation of p53 in cells expressing the Y187F/C189A and T193A mutants. IFN/RA had a slight stimulatory effect on the -ID, and S199A/Y200F-induced DNA binding of p53. These data are consistent with the luciferase expression data.

View larger version (53K): [in this window] [in a new window]   Fig. 9.   Effect of IFN/RA on cells expressing TR mutants. A, cells stably expressing FLAG-tagged TR mutants were selected for 3 weeks with IFN- (500 units/ml) and RA (1 µM). Cells were then stained with sulforhodamine B to visualize surviving cells. B and C, expression of TR mutants in the stable transfectants. Extracts from the cells transfected with the indicated plasmids (70 µg) were employed for Western blot analysis using anti-FLAG antibodies.

To examine the role of these mutants in IFN/RA-induced cytotoxicity, stable transfectants expressing the FLAG-tagged mutants were plated and selected with IFN/RA and G418 (0.5 mg/ml) for 3 weeks. The Y187F/C189A and T193A mutants resisted IFN/RA, unlike TR, -ID, and S199A/Y200F, which were killed by the combination (Fig. 9A). This property is consistent with their respective p53-augmenting functions. Expression of these mutants in the cells was confirmed by a Western blot analysis with FLAG tag-specific antibodies (Fig. 9, B and C). However, the expression of S199A/Y200F was lost over several passages, indicating its growth-inhibitory effect.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Trx, a ubiquitous redox protein, regulates a wide array of cellular activities including growth, transcription, and immune responses (23). Its redox status is controlled by a cytosolic enzyme, TR. In mammals, three different TR genes, TR1, TR2, and TR3, which are expressed in an organelle- and tissue-specific manner, have been identified to date (39, 40). The founding member of this family, TR1, is expressed ubiquitously (23). Physiologic roles of the new members of TR family are unclear at present. Mammalian TR has broader substrate specificity than its prokaryotic homologues (23). For example, it can reduce unrelated compounds such as selenite, alloxan, 5,5'-dithiobis(2-nitorbenzoic acid), vitamin K, selenocysteine, selinodiglutathione, and S-nitrosoglutathione, in addition to Trx. Because TR modifies its substrates extremely rapidly, and no stable prosthetic groups are involved in this process (23), it has been technically difficult to define its intracellular targets. Trx is also implicated in maintaining the redox status of transcription factors, AP1, PEBP2, nuclear factor B, and HIF1 (28, 41-44).

Based on chemical inhibition and other correlative data, Trx1 is implicated in cell growth promotion in some cell types (24). However, it is not a universal phenomenon. Several studies showed that Trx and TR participate in growth-inhibitory actions. The Trx homologue of Drosophila, deadhead, is essential for female meiosis and early embryonic development (45) but not for DNA synthesis in vivo (46). Deletion of TR1 in Drosophila affects survival (47). It codes for a cytoplasmic and a mitochondrial isoform, both of which are necessary for the viability of the fly. Similarly, Trx1-null mouse embryos do not implant at all (48). Thioredoxin inhibits DNA synthesis in the fertilized Xenopus eggs (49). In yeast, deletion of Trx gene causes an increase in the frequency of mitotic cell cycles (50). TR is necessary for p53-dependent growth suppression in fission and budding yeasts (51, 52). Inhibition of Trx1 causes resistance to IFN--induced apoptosis (53), and Trx-producing hepatomas grow slowly (54). Using a genetic tool, we have previously shown that TR1 is critical for the cell death controlled by the IFN/RA combination (22, 55). We and others have shown that tumor suppressor p53, caspase-8, and caspase-3 are also regulated by TR:Trx (25-27, 56, 57).

Tumor suppressor p53 inhibits cell growth either by aborting the division cycle or by inducing death (58) and is a frequently inactivated gene in several human cancers. Its activity is controlled by several posttranslational mechanisms such as stability, phosphorylation by protein kinases, ubiquitination, SUMOylation, acetylation, and redox factors (59). We have reported previously that the IFN/RA combination promotes p53-dependent gene expression and cell death using TR1 and Trx1 (27) without elevating the levels of p53 protein. p53 has two cysteines (residues 242 and 176), which, along with two histidines, coordinate Zn²⁺ for DNA binding (60, 61). Oxidation of these cysteines ablates its transcriptional activity (60, 62). Wild-type human p53, but not a cysteine mutant, inhibits growth in yeast (63, 64). Redox control of p53 activity is further substantiated by studies that showed that tms1, a dehydrogenase, suppresses p53-induced growth arrest (65). Lastly, studies using chimeric p53 proteins in yeast have revealed that in addition to the DNA binding site, the transactivation domain of p53 is also subject to TR-dependent redox regulation (66). Retinoids are known to activate oxidative stress through an increased synthesis of reactive oxygen species (67-70). The presence of ROS is a signal for p53 activation (58). IFN/RA treatment induces TR and Trx in cells undergoing apoptosis (22, 26). p53 is kept in a reduced form by redox factor-1, a downstream effector of the TR:Trx system. Indeed, recent studies have shown that redox factor-1 promotes p53-dependent gene expression and cell death (33). Redox factor-1 has been suggested to prevent the oxidation of cysteine residues of p53. Trx1 augments redox factor-1-dependent gene expression through p53 (57). More importantly, the failure of -ID to promote gene expression in the absence of p53 (Fig. 5) clearly indicates an obligatory role for p53 in this process. Because -ID fails to induce the p53-responsive reporter in cells expressing a mutant Trx, its effects are Trx-dependent (Fig. 5).

The present observations raise a question: what is the role of ID? Mammalian TR is a dimeric selenoprotein. Mutational analyses have revealed that the mammalian TR has two redox centers: one at the N terminus in the FAD binding domain, and the other at its C terminus. The phylogenetically well-conserved N-terminal redox center is constituted by cysteines residues at positions 59 and 64. A selenocysteine at the C terminus forms the C-terminal redox center. In an unusual manner, one of the two in-frame stop codons at the 3' end of the open reading frame in conjunction with a stem-loop structure formed by a selenocystine incorporation sequence of the 3'-untranslated region of TR mRNA is proposed to act as an acceptor for selenocysteinyl tRNA, leading to a co-translational incorporation of selenocysteine into the mature protein (24, 71). Recent studies showed that bovine and rat TRs lacking this selenocysteine have an extremely reduced k_cat value in vitro (72, 73). However, this enzyme did not completely lose its enzymatic activity. Furthermore, depending on the substrate used for the assay, its activity is normal as long as the N-terminal primary redox center is retained (73). In fact, incubation of the mutant-derived protein with selenocysteine restored the activity dramatically in vitro (72). These data suggest that selenocysteine in trans can restore the enzyme activity. However, a low occurrence of free selenocysteine (74) suggests that such modulation is a less frequent event in vivo. It is possible that stress conditions such as apoptosis alter the physiologic availability of selenocysteine. Such selenocysteine may be derived from apoptotic degradation of other selenoproteins or its biosynthesis. At least 10-12 proteins that contain selenocysteine (75) have been identified to date, and the degradation of these proteins may provide free selenocysteine.

The observations that yeast TR, which lacks a selenocysteine (76, 77), can promote p53-dependent cell growth arrest (51, 52) and that a truncated human TR lacking its ID also promotes p53-dependent transcriptional response (this study) are consistent with the proposition that TR-stimulated p53-mediated responses are selenocysteine-independent. Similarly, Drosophila and plant TRs do not require selenium for their function (47, 78). In fact, plant TR behaves like a prokaryotic TR (78). Furthermore, the prokaryotic TRs do not reduce other substrates, like the mammalian TR. Recent crystallographic data on rat TR (79) have shown that the enzyme forms a head-to-tail dimer, with the N-terminal redox center buried inside. Reducing power is first transferred from NADPH to the N-terminal redox center. The C terminus of one subunit is inserted into the charged cleft (N-terminal redox center) of the other subunit to tap electrons from the active site to selenocysteine. The reduced selenocysteine then donates electrons to Trx/other substrates at the surface of TR dimer, after emerging from the catalytic cleft. In this model, the ID extends like a robotic arm to transfer electrons between the enzyme and its various substrates. In the case of prokaryotic TR, which also acts as a dimer, after the receipt of electrons at the redox center from NADPH, the NADPH binding domain undergoes a 66° rotation to provide access to oxidized Trx to the redox center (23, 24). -ID may behave like prokaryotic, yeast, and plant TRs with strict Trx-reductive properties. This would be sufficient for activating p53. Thus, it would appear that a selenocysteine at the C terminus has evolved to enhance catalysis and broaden substrate range and is an optional accessory. Although an augmentation of TR activity by selenium has been reported in mammalian cells, these studies used a supranutritional concentration of selenium and are not physiologically relevant. In fact, an inverse correlation between selenium and TR activity has also been reported (80). Lastly, selenium metabolites can activate cell cycle arrest in the absence of p53 and DNA damage (80), indicating the existence of a separate mechanism of action. In light of these data, we suggest that selenium/selenocysteine plays a limited role in TR-mediated growth-suppressive pathways in vivo but is required for other redox reactions during normal growth.

Alternatively, TR bearing alkylated C-terminal selenocysteine and cysteine residues exhibits 30-fold higher NADPH oxidase activity compared with the wild-type enzyme and is capable of producing superoxides (81). These superoxides can oxidize intracellular environment, thus tilting the balance toward p53 activation and cell death. Consistent with this suggestion, deletion of the NADPH binding domain (mutants -ID60 and -ID₅₀) prevents the stimulatory effects of TR on p53-dependent gene expression (Fig. 7). It is interesting to note that the mutation of cysteine residue at 189 depletes the p53-augmenting function of -ID (Fig. 8). This suggests that Trx transiently interacts with this site during enzymatic modification because Trx1 is still necessary to promote p53-dependent gene expression. The threonine and tyrosine residues may undergo posttranslational modifications, which in turn contribute to full activity of -ID. Thus, ID, which is unique to mammalian TR, appears to act as a regulatory switch in cell growth control. Its presence attenuates the growth-inhibitory effect of TR, and its removal promotes cell death. Because prokaryotic/unicellular organisms do not undergo apoptosis, and the primary role of TR is only to maintain the redox functions. Thus, these organisms have a much simpler modular structure. Because TR and Trx can promote pro- and anti-growth processes, depending on the physiologic status of cells, and because mammalian TR has expanded physiologic roles, the evolution of a regulatory switch (ID) is critical for preventing an inadvertent operation of these divergent processes. Thus, ID may act as a decision switch to mediate such "yin-yang" reactions in vivo. In the meantime, one potential use for the -ID mutant will be in gene therapy along with wild-type p53 in p53-null tumors.

	ACKNOWLEDGEMENT

We thank Peter Gutierrez for helpful discussions.

	FOOTNOTES

^* This work was supported by National Cancer Institute Grants CA 78282 and CA 71401 (to D. V. K.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Both authors contributed equally to this work.

^§ Present address: Dept. of Surgery, Tongji Medical Center, Wuhan, People's Republic of China.

To whom correspondence should be addressed: Greenebaum Cancer Center, University of Maryland School of Medicine, 665 W. Baltimore St., BRB-9^th Floor, Baltimore, MD 21201. Tel.: 410-328-1396; Fax: 410-328-1397; E-mail: dkalvako@umaryland.edu.

Published, JBC Papers in Press, April 12, 2002, DOI 10.1074/jbc.M202286200

	ABBREVIATIONS

The abbreviations used are: IFN, interferon; RA, all-trans-retinoic acid; STAT, signal transducers and activators of transcription; TR, thioredoxin reductase; Trx, thioredoxin; NBD, NADPH binding domain; ID, interface domain; CMV, cytomegalovirus; Luc, luciferase; FITC, fluorescein isothiocyanate; EMSA, electrophoretic mobility shift analysis.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES