A Polymorphism in the 3' Untranslated Region of the Gene for Tumor Necrosis Factor Receptor 2 Modulates Reporter Gene Expression
http://www.100md.com
内分泌学杂志 2005年第5期
Department de Fisiologia (I.P., B.L., M.Bu., E.E.), Facultat de Biologia, Universitat de Barcelona, E-08028 Barcelona, Spain; Diabetes, Endocrinology, and Nutrition Research Unit (B.L., J.M.F.-R.), University of Girona, Hospital of Girona "Dr Josep Trueta," E-17007 Girona, Spain; and Universitat Rovira-Virgili (M.Br., J.V.), 43003 Tarragona, Spain
Address all correspondence and requests for reprints to: Dr. Enric Espel, Department Fisiologia, Facultat Biologia, Diagonal 645, 08028 Barcelona, Spain. E-mail: eespel@ub.edu.
Abstract
The gene encoding the human TNF receptor (TNFR) 2 contains polymorphisms in the 3' untranslated region (UTR). Previous studies have shown that some variant alleles in this region are associated with obesity and insulin resistance. However, the effect of these polymorphisms on the expression of TNFR2 has not been studied to date. To examine the role played by different haplotypes in the control of TNFR2 expression (haplotypes A1-A5, referring to nucleotides 1663 G/A, 1668 T/G, and 1690 T/C), we introduced these sequences into the 3'-UTR of a heterologous reporter gene and expressed the corresponding constructs in a human T-cell line. We demonstrate that a 485-nt fragment of the TNFR2 3'-UTR that contains a U-rich region decreases reporter expression and that haplotypes A1-A4 exert a stronger effect than A5. Furthermore, time-course assays of mRNA stability using actinomycin D revealed that haplotypes A1-A4 destabilize the mRNA. The proximal TNFR2 3'-UTR, independently of haplotype differences, responded to T-cell activation by increasing mRNA decay. Electromobility shift analysis demonstrated that protein(s) found in T-cell extracts bind to the 485-nt fragment. We suggest that an increased rate of TNFR2 mRNA decay protects cells from unrestrained TNF effects and that this protection is weakened in A5 subjects. These findings may explain the association of this haplotype with obesity and increased leptin levels.
Introduction
TNF IS A CYTOKINE THAT plays a central role in inflammatory phenomena and is involved in the pathogenesis of chronic local or systemic inflammation in vivo (1). The effects of TNF are signaled by two cell surface receptors, TNFR1 (TNFRSF1A or p55) and TNFR2 (TNFRSF1B or p75/80), which mediate cellular responses ranging from apoptosis to cell proliferation and differentiation (2). TNFR2 is predominantly expressed in cells of the hematopoietic lineage and is the principal TNFR expressed by blood mononuclear cells and activated T cells (2, 3, 4), playing a key role in T-cell homeostasis (5, 6, 7, 8, 9, 10).
These two receptors are also produced in a soluble form. This form constitutes the extracellular portion of membrane-associated TNFRs, shed by proteolytic cleavage (11, 12). Soluble TNFR2 (sTNFR2) inhibits TNF action by competing with cell surface receptors to bind TNF. Chronic enhanced production of sTNFR2 is associated with several human inflammatory and autoimmune conditions, including sepsis (13), lupus (14), and rheumatoid arthritis (15). A ligand-independent role for this receptor has also been reported. Thus, transgenic mice that constitutively express enhanced levels of a wild-type human TNFR2 suffer from a severe multiorgan inflammatory syndrome (16).
The role of the TNF-TNFR pathway in insulin resistance and lipid abnormalities has been extensively described (17, 18). Mice with targeted null mutations in the genes that encode TNFR1 and 2 show a significant increase in insulin sensitivity (19), which indicates a primary role for the TNF-TNFR pathway in the development of insulin resistance. Levels of RNA transcripts encoding TNFR2 receptors are significantly increased in the fat and muscle of obese diabetic animals (20). The TNF-TNFR system may also contribute to the regulation of insulin action in human obesity (18). In obese humans and insulin-resistant and diabetic subjects, production of TNF in muscle and adipose tissue is higher than in lean individuals (21, 22, 23), a condition that can be reversed by weight loss (21, 22). Furthermore, plasma levels of sTNFR2 are correlated with basal and postload glucose levels and insulin resistance (24, 25).
The 3'-untranslated region (UTR) of human TNFR2 shows a degree of polymorphism (26). An increase in the frequency of some alleles has been correlated with insulin resistance (27) and Crohn’s disease (28). Interestingly, other mutations in this locus have been associated with several components of insulin resistance syndrome, such as hypertension (29), dyslipidemia (30), and cardiovascular disease (31), and with degenerative diseases such as osteoporosis (32). This locus has also been linked to polycystic ovary syndrome, a condition with known defects in insulin activity (33).
To date, no function has been ascribed to the polymorphisms present in the 3'-UTR of TNFR2. To determine whether these polymorphisms affect gene expression, we fused a genomic DNA fragment containing the polymorphic region of haplotypes A1-A5 to a green fluorescent protein (GFP) reporter gene. We then used flow cytometry to study the expression of this gene in live Jurkat T cells. T cells represent an interesting model with which to study the control of TNFR2 expression because these cells are known to regulate TNFR2 levels on activation (3, 6). This means that the biochemical machinery necessary to regulate TNFR2 expression is present in these cells. Therefore, we decided to undertake the present study in lymphoid Jurkat T cells. Our results indicate that haplotypes A1-A4 decrease the expression of TNFR2 (p75/80) by increasing the decay rate of its mRNA.
Materials and Methods
Chemicals and antibodies
DMEM, penicillin, streptomycin, and fetal calf serum (FCS) were purchased from Sigma-Aldrich (Madrid, Spain). FCS was decomplemented at 56 C for 30 min. Actinomycin D (ActD) was purchased from Tocris (Biogen, Madrid, Spain). Anti-GFP was obtained from Roche Diagnostics (Barcelona, Spain), and antihuman actin was from Sigma-Aldrich. Anti-CD (aCD) 3, a mouse IgG2a monoclonal antibody (33–2A3 workshop II; a kind gift from Dr. R. Vilella, Hospital Clínic, Barcelona, Spain), was used as a hybridoma supernatant at a dilution of 1:1000. This dilution was shown to induce maximal TNF secretion in Jurkat T cells (data not shown). aCD28, a mouse IgM monoclonal antibody (CK243; a kind gift from Dr. Pedro Romero, Ludwig Institute for Cancer Research, Lausanne Branch, Switzerland), was used as a hybridoma supernatant at a dilution of 1:20. This dilution was shown to produce maximal TNF secretion in combination with aCD3 in Jurkat T cells (data not shown).
Reporter plasmid construction
A vector containing an enhanced GFP coding sequence flanked by ?-globin UTRs and directed by the cytomegalovirus (CMV) promoter (pCS3glob-GFP-glob; see below) was used to analyze the putative regulatory sequences of the TNFR2 gene. TNFR2 3'-UTR genomic DNA fragments were introduced into pCS3glob-GFP-glob downstream of the GFP coding region and upstream of the ?-globin 3'-UTR.
To generate a reporter construct that contained haplotypes A3 and A5 in a sense orientation and haplotype A2 in an antisense orientation, genomic DNA was isolated from homozygous donors for each TNFR2 haplotype. A 485-bp fragment of the TNFR2 3'-UTR from A2, A3, and A5 donors was generated by PCR using the following primers: 5'-GGGGAGCACCGAAGAGAA and 5'-AAGGATGCTGGGGTTTTCT. The PCR product was cloned between the GFP gene and the globin 3'-UTR (into the EcoRV site) in the vector pCS3glob-GFP-glob. The construct containing haplotype A3 was used to prepare plasmids containing haplotypes A1, A2, and A4 by site-directed mutagenesis. In this way, we obtained haplotypes A1-A5 with the following variations at positions 1663 A/G, 1668 T/G, and 1690 T/C (Table 1), which spread from the stop codon (TAA, Fig. 1) some 188, 193, and 215 nt downstream, respectively (National Center for Biotechnology Information, NM 001066).
TABLE 1. Characterization of variants in the 3'-UTR (exon 10) of the TNFR2 gene and frequency of these variants in the population
FIG. 1. Linearity of response of the GFP reporter in Jurkat T cells. A, Schematic diagram of the GFP reporter gene, indicating the coding region and the ?-globin 3'-UTR fragment. Constructs are based on the vector pCS3+mT (35 ). The TNFR2 haplotypes A1-A5 (the polymorphic nucleotides are indicated) and A2 in antisense orientation (A2as) were assembled downstream of the GFP coding region, as shown. B, Example of FACS analysis with GFP-A5-transfected cells. Histogram showing the fluorescence of cells transfected with 20 μg of plasmid GFP-A5 or with empty vector pCS3+mT (without GFP; dark area). The relative amount of emitted fluorescence from each cell is shown on the horizontal axis. The data were grouped into four intervals, as indicated by the top bar. From these data, the MFI of the sample was obtained by multiplying the percentage of cells within each interval by its MFI and adding the result of the four intervals (36 ). C, Relative expression of CMV-GFP in Jurkat T cells. Cells (1.0 x 107) were transfected with increasing amounts of the CMV-GFP vector along with 5 μg of a luciferase-expressing vector (pGL2). Twenty-four hours later, GFP expression was analyzed in live cells by FACS. The MFI normalized by the transfection efficiency (luciferase) is shown on the y-axis, and the amount of DNA transfected on the x-axis. MIF was calculated as described above (B). Data represent the average of three independent transfections ± SD. D, The sequence of the TNFR2 3'-UTR fragment (485 bp) for haplotype A2 is shown with the position of nine U-rich stretches containing three or more Us underlined. The U-rich cluster is reminiscent of AREs. The three polymorphic nucleotides are shown in lowercase and circled. The 485-bp fragment was inserted into a unique EcoRV restriction site localized between GFP and the globin 3'-UTR (A).
The constructs pCS3glob-GFP-glob and pCS3glob-GFP-A3glob were fully sequenced (ABI PRISM, PerkinElmer, Barcelona, Spain) before use in transfection experiments. Polymorphism 1663 A/G was also controlled by FauI digestion. The rest of the plasmids were sequenced through the GFP-coding and 3'-UTR region. For simplicity, the plasmids are called CMV-GFP (pCS3glob-GFP-glob), GFP-A1 (pCS3glob-GFP-A1-glob), GFP-A2 (pCS3glob-GFP-A2-glob), GFP-A3 (pCS3glob-GFP-A3-glob), GFP-A4 (pCS3glob-GFP-A4-glob), GFP-A5 (pCS3glob-GFP-A5-glob), and GFP-A2as (A2 in antisense orientation, CS3glob-GFP-A2as-glob).
Plasmid CMV-GFP was generated as follows: plasmid pT7globGFPglob, which contains xenopus ?-globin 5'- and 3'-UTRs flanking the GFP sequence, was a gift from Dr. A. A. M. Thomas (Department of Developmental Biology, University of Utrecht, Utrecht, The Netherlands) (34). The fragment globGFPglob was excised by digestion with HindIII and EcoRI and then subcloned into the vector pCS3+mT downstream of the CMV promoter (35).
Transfections
Jurkat T cells were grown in DMEM containing 10% FCS. Cells were diluted to 0.5 x 106 cells/ml the day before transfection. The next day they were pelleted, washed in PBS, counted, and resuspended at a final concentration of 5 x 107 cells/ml in 0.2-cm-wide electroporation cuvettes (Invitrogen, Barcelona, Spain). Transient transfection was performed by electroporation. Briefly, 10 μg (or the amount indicated in figure legends) of GFP constructs and, where indicated, 5 μg of the luciferase reporter plasmid pGL2 (Promega, Innogenetics, Barcelona, Spain) were transfected into cells in a 200 μl cell suspension (1700 μF, 72, 126 V, 30 msec, using a BTX 600). Cells were then seeded into petri dishes at a final density of 5 x 105 cells/ml in DMEM-10% FCS. Experiments were carried out 24 h after transfection. Cells (0.5 x 106 in 1 ml) were taken directly from culture flasks and placed in 3-ml plastic tubes for fluorescence-activated cell sorting (FACS) analysis. One microliter of propidium iodide (0.07%) was then added, and tubes were maintained on ice in the dark before flow cytometry studies. Luciferase activity was measured in cell extracts according to the manufacturer’s instructions (Promega, Innogenetics). The transfection efficiency of most experiments was determined from the percentage of cells displaying green fluorescence (36). This procedure provided more consistent results than cotransfecting luciferase. Transfection efficiency values varied between 5 and 70%.
Flow cytometry
Transgene expression in live cells was measured by flow cytometric analysis on an Epics XL flow cytometer (Coulter Corp., Miami, FL) equipped with a 488-nm argon laser at 15 mW. Analysis was performed according to the method described by Kar-Roy et al. (36). This method offers advantages over traditional luciferase or ?-galactosidase reporter genes in that GFP transgene expression can be analyzed in individual, live-cell populations. The instrument was set up with the standard configuration: forward scatter, side scatter, and green (525 nm) fluorescence for GFP were collected. Green fluorescence was represented on a logarithmic histogram. The relationship between the fluorescence signal and the number of emitting chromophores per cell is linear, a higher signal indicating higher expression levels of the reporter (36).
Twenty-four hours after transfection, the cells (0.5 ml, containing approximately 2.5 x 105 cells) were analyzed. Gating for live cells was achieved by forward scatter and side scatter. The histogram of the intensity of green fluorescence in the live cells was used to analyze GFP expression. Quantification of 3'-UTR regulatory effects on GFP expression was performed by measuring the median of the fluorescence intensity histogram obtained from the transfected population (36). Differences in GFP expression, measured by the median fluorescence intensity (MFI), indicate distinct posttranscriptional regulatory activity of the UTRs tested. Because this system was based on a transient transfection assay, the population of transfected cells was heterogeneous in terms of their GFP expression levels (36). Optical alignment was based on optimized signal from 10 nm fluorescent beads (Flowcheck, Epics Division, Coulter Corp.). Time vs. fluorescence was used as a control for the stability of the instrument. GFP expression results are presented as MFI/transfection efficiency.
Western blotting
Cells (1.0 x 106) were lysed in 100 μl Laemmli sample buffer. The samples were heated at 95 C for 30 min and separated by SDS-PAGE. A 15% gel was used for GFP analysis. The proteins were blotted onto polyvinylidene difluoride membranes and detected by immunoblot. GFP was detected with a mixture of mouse monoclonal antibodies (Roche Diagnostics) at a dilution of 3 μg per 8 ml.
After stripping of anti-GFP (Re-blot plus; Chemicon, Pacisa-Giralt, Madrid, Spain), the same membrane was reprobed with an antiactin polyclonal antibody raised in rabbit, diluted at 1:200 (reference A-2066; Sigma-Aldrich).
Ribonuclease (RNase) protection assay and GFP probes
RNase protection assays (RPA) were performed as described previously (37). Digested RNA was resolved on a 6% polyacrylamide-urea denaturing gel. In some experiments, RPA analysis was performed directly in cell lysate, as described elsewhere. In these cases, digested RNA was fractionated on a 6% polyacrylamide/Tris-borate-EDTA nondenaturing gel and vacuum dried onto nylon membranes (Nytran Supercharge; Sleicher & Schuell, Acefesa, Barcelona, Spain). Radioactivity of protected bands was measured by phosphor imager analysis using a FX molecular imager (Bio-Rad Laboratories, Barcelona, Spain).
To prepare the GFP probe, a GFP DNA fragment was cloned into the pSTBlue-1 vector (Novagen, Bionova, Madrid, Spain). The recombinant plasmid was digested with FokI. T7 polymerase generates an antisense RNA of 494 nt from the FokI-digested plasmid. RNase digestion of samples that contain glob-GFP-glob mRNA hybridized to the antisense RNA probe generates a protected probe fragment of 338 nt.
The probe used in the RPA (see Fig. 4) was prepared as follows: a DNA fragment of 254 bp was generated by PCR from the plasmid CMV-GFP using the primers GFP forward AGCAAGGGCGAGGAG and T7-GFP reverse GAATTCTAATACGACTCACTATAGGTGTGGTCGGGGTAGC. T7 polymerase was then used to generate an antisense RNA of 231 nt from that fragment. The protected fragment after hybridization with GFP RNA and digestion with RNases A and T1 is 229 nt in length, a 2-nt difference from the nondigested one.
FIG. 4. Stimulation of Jurkat cells does not reverse the inhibition of GFP expression by the TNFR2 3'-UTR. A, Jurkat cells were transfected with GFP-TNFR2 constructs, as indicated in Fig. 2. Cells were starved by culturing in DMEM-0.1% FCS. Twenty hours later, cells were stimulated with aCD3 and aCD28 antibodies for 5 h, and GFP expression was analyzed by FACS. Results are shown as fold increase over control and nonstimulated cells. Experiments were performed in duplicate; the data shown are the mean values of at least three independent transfections ± SD. B, C, and D, Jurkat cells were transfected with CMV-GFP and stimulated with aCD3/aCD28 antibodies for 1 h, and GFP expression was analyzed by FACS (B) and RPA (C and D). The labeled antisense RNA probe used for GFP in this RPA was 231 nt long, generated as indicated in Materials and Methods. Arrowheads (D) indicate the positions of the undigested probes. Arrows indicate the positions of protected fragments. E, Nuclear run-on analysis of GFP gene activity in transfected cells. Total [32P]RNA isolated from Jurkat T cells transfected with plasmid GFP-A2 or CMV-GFP was hybridized to slot blots containing TNFR2 3'-UTR DNA or GFP DNA, as described in Materials and Methods. Bars, obtained by phosphor imaging of slot blots, represent the relative transcription rates of GFP-A2 and CMV-GFP genes. Data represent the mean of two independent experiments.
GFP mRNA decay
For the GFP mRNA decay experiments, transfected cells were incubated for 12 h at 37 C in DMEM-0.2% FCS. aCD3 and aCD28 monoclonal antibodies were added and cells were incubated at 37 C for a further 3 h. ActD (8 μg/ml) was added to the culture dishes and cells incubated for varying lengths of time, up to 3 h. Cells were pelleted by centrifugation and lysed with 0.4 ml of 4 M guanidium thiocyanate, 25 mM Na-citrate (pH 7.0), 0.5% N-lauryl sarcosine. RPA analysis was performed directly in cell lysates, as described elsewhere (38), using the GFP and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probes generated with T7 RNA polymerase (see RNase protection assay). Radioactivity of protected bands was measured by phosphor imager analysis. The phosphor imager values were represented vs. time with ActD, and the slopes were analyzed by linear regression analysis.
RNA EMSA
Jurkat cell extracts were prepared as described previously (39, 40). Briefly, cells (1.0 x 108) were pelleted, washed with PBS, resuspended in 1 ml PBS, and centrifuged at 13,000 x g for 10 sec in Eppendorf tubes to remove the supernatant. The cell pellet was lysed in 830 μl Munro-lysis buffer [10 mM HEPES (pH 7.6), 3 mM MgCl2, 40 mM KCl, 5% glycerol, 1 mM dithiothreitol, leupeptin, aprotinin] supplemented with 0.2% Nonidet P-40 by pipetting 10 times. The cell lysate was diluted with 1.66 ml of Munro-lysis buffer and centrifuged at 13,000 x g for 2 min to pellet the nuclei. The supernatant (cell extract) was aliquoted and kept frozen at –80 C until used. To prepare RNA probes, the 485-bp TNFR2 3'-UTR fragment was amplified by PCR downstream of the T7 promoter using the oligonucleotides forward GGTAATACGACTCACTATA-GGGGAGCACCGAAGAGAA (T7 promoter linked to TNFR2 3'-UTR) and reverse AAGGATGCTGGGGTTTTCT, using the GFPA1-A5 constructs as templates. The amplified fragments were purified (QiaQuick; QIAGEN, Izasa, Barcelona, Spain) and incubated with T7 RNA polymerase to generate 32P-uridine-labeled RNA probes. Assays consisted of 15 μg of Jurkat cytosolic extract incubated with 50,000 cpm of labeled RNA, 4 μg/μl yeast tRNA, and 2 μg/μl heparin in a total volume of 17 μl. Reactions were incubated at room temperature for 15 min, followed by 10 min with RNase T1 (150 U). Samples were resolved on 4% nondenaturing polyacrylamide gels.
Run-on
To monitor transcriptional activity, nuclear run-on assays were performed as described elsewhere (41). Nonradioactive probes for GFP and TNFR2 were prepared by PCR using the following oligonucleotides: GFP, 5'-AGCAGCACGACTTCTTCA and 5'-CGTTGTGGCTGTTGTATT; TNFR2, 5'-GGGGAGCACCGAAGAGAA and 5'-AAGATGCTGGGGTTTTCT. RNA labeling in run-on transcription mixtures containing 6.0 x 107 nuclei from transfected cells was performed using [-32P]UTP. RNA was extracted with phenol/chloroform/isoamyl alcohol and precipitated with ethanol. Nylon membranes (NYTRAN SuPerCharge; Schleicher & Schüell) containing the nonradioactive probes were hybridized to the labeled RNA at 68 C for 72 h. After hybridization, the membranes were washed once with 2x saline sodium citrate (SSC), 0.1% sodium dodecyl sulfate (SDS) for 5 min at room temperature, twice with 0.5x SSC, 0.1% SDS for 20 min at 68 C, and finally, once with 0.1x SSC, 0.1% SDS for 20 min at 68 C. The radioactivity contained in the bands was determined using a molecular imager (Bio-Rad).
Analysis of RNA secondary structure
RNA secondary structure analysis was performed using the MFOLD program designed by Mathews et al. (42, 43) (available at http://www.bioinfo.rpi.edu/applications/mfold). The folding temperature was fixed at 37 C.
Statistical analysis
Data are expressed as the mean ± SD. Statistically significant differences between samples were determined using a one-factor ANOVA followed by a t test, except for flow cytometry data for which the Bonferroni multiple comparisons test was used.
Results
To evaluate a potential functional role for the polymorphisms located in the 3'-UTR of TNFR2 (Table 1), we introduced a 485-bp sequence that contained the polymorphic region of the TNFR2 3'-UTR into reporter constructs (Fig. 1). TNFR2 sequences were introduced between the GFP gene, used as a quantitative reporter (36, 44, 45, 46), and the stable ?-globin UTR. Because the fluorescence of transiently transfected cells is heterogeneous, the MFI of each sample was calculated by dividing the histogram into four quadrants and determining the fluorescence for each quadrant (Fig. 1B) (36). The fluorescence intensity emitted by cells transfected with constructs containing the GFP gene is proportional to the amount of GFP produced by the transfected cells and increased progressively as a function of the amount of DNA transfected (Fig. 1C), thereby confirming the validity of the procedure.
Haplotypes A1-A4 reduce expression of a GFP reporter
The fluorescence of transfected cells displayed a linear increase with increasing amounts of the different constructs (Fig. 2A). However, insertion of the TNFR2 haplotypes A1, A2, A3, or A4 similarly reduced the expression of GFP to 50% of that observed with CMV-GFP at any of the three DNA concentrations tested (Fig. 2A; P < 0.05, n = 6). The reduction in the expression of GFP by haplotype A5 was less pronounced (35% at 20 μg; P < 0.05, n = 8). Furthermore, expression of GFP-A5 was significantly higher than any of the other haplotypes A1-A4 (observed at 20 μg, P < 0.05, n = 8). As a control, A2 used in an antisense orientation (A2as), displayed similar levels of GFP expression to those found with CMV-GFP.
FIG. 2. Expression of GFP-3'-UTR-TNFR2 in Jurkat T cells. A, Regulation of GFP expression by TNFR2 3'-UTR haplotypes. Cells (1.0 x 107) were transfected with increasing amounts of plasmids containing haplotypes A1-A5. Twenty-four hours later, fluorescence was analyzed in live cells by FACS. The MFI for GFP was calculated and normalized to the transfection efficiency. The normalized ratio (GFP fluorescence to transfection efficiency) is represented as a percent of the maximum value. Measurements were performed in duplicate; the data represent the mean ± SE from at least six independent experiments. *, Significant differences (P < 0.05, n = 6). B, Transfection efficiency. Transfection efficiency in A was determined from the percentage of GFP-positive (+ve) cells in each sample (36 ). The data shown are the mean values of at least six independent experiments ± SD. C, Analysis of GFP expression by Western blot. Cells transfected in A with 20 μg of the indicated constructs were lysed and proteins fractionated by SDS-PAGE (15% gels) and transferred to a membrane. GFP expression was detected with an anti-GFP monoclonal antibody. The same membrane was reprobed with an antiactin antibody. Apparent molecular weight is indicated on the right of the figure. This is a representative experiment from a total of three. D, Analysis of GFP mRNA by RPA. Total RNA was prepared from cells transfected in A and GFP mRNA analyzed by RPA with probes for GFP and cellular GAPDH. Protected RNAs were fractionated on a urea denaturing gel and detected by autoradiography. Arrowheads indicate the position of undigested probes. Arrows indicate the position of protected fragments. Top, Film was overexposed to show differences in GFP expression between samples. P, Probes; Pr, probes treated in the same way as the rest of the samples but without RNases; Pd, probes digested with RNases as for the rest of the samples; –, cells transfected with empty-vector (lacking GFP). The experiment was repeated three times and a representative result is shown. E, GFP mRNA expression in transfected cells. Radioactivity of protected bands in D was measured using a Bio-Rad molecular phosphor imager and the GFP to GAPDH ratio calculated to normalize differences in sample loading. In these experiments, the transfection efficiency was calculated by cotransfection of a luciferase reporter (pGL2). The normalized GFP to GAPDH to luciferase (Luc) ratio is shown as a percent of the maximum value. Data shown are mean values ± SD from three independent experiments. A3 expression differed from that of CMV-GFP (*, P < 0.05). F, Analysis of GFP mRNA by RPA. RPA analysis was performed directly in guanidium thiocyanate lysates of cells transfected in A. Digested RNA was resolved on a 6% nondenaturing polyacrylamide gel, which was then vacuum dried and analyzed by phosphor imaging. Arrows indicate the position of protected fragments. A1-A4, Cells transfected with individual reporter plasmids containing haplotypes A1, A2, A3, or A4 mixed before lysis with guanidium thiocyanate. The experiment was repeated four times, and a representative result is shown. G, GFP mRNA expression in transfected cells. Radioactivity in protected bands from F and similar experiments was measured by phosphor imaging, and the GFP to GAPDH ratio calculated to normalize differences in sample loading. Transfection efficiency was determined from the percentage of GFP +ve cells. The normalized ratio (GFP to GAPDH) to transfection efficiency is shown as a percent of the maximum value. The data are shown as mean values ± SD. A1-A4 expression differed from CMV-GFP, *, Differences statistically significant in nonstimulated (P < 0.01) (DMEM, n = 5) and stimulated conditions (cells stimulated with aCD3+aCD28 antibodies, n = 4).
The amount of the GFP protein in transfected cells was also analyzed by Western blot (Fig. 2C). Gel loading was verified by reprobing with an antiactin antibody. Although loading of cell lysates from samples transfected with GFP constructs containing A1 to A5 or CMV-GFP was similar, the amount of GFP in cells transfected with GFP-A1 to GFP-A4 was lower, confirming the results obtained by analysis of GFP fluorescence.
Haplotypes A1-A4 reduce the expression of GFP mRNA
Because the differences between haplotypes A1-A5 and CMV-GFP reside in the 3'-UTR, the results in Fig. 2A suggest that the observed differences in GFP expression correspond to posttranscriptional mechanisms. This could correspond to differences in mRNA decay and/or mRNA translation. To ascertain which of the two mechanisms was causing the difference in expressed GFP, the level of GFP mRNA was compared between transfectants. Because the four haplotypes A1-A4 behaved similarly in controlling the expression of GFP (Fig. 2, A and C), RNA analysis was performed on cells transfected with GFP-A3 and in a mixture of cells transfected with each of the constructs GFP-A1, GFP-A2, GFP-A3, and GFP-A4. As expected, haplotype A2 in antisense orientation did not modify the GFP mRNA steady-state levels observed when transfecting control CMV-GFP (Fig. 2, D and E). However, the expression of GFP-A3 mRNA followed a pattern similar to that of the GFP protein, with a lower accumulation of this mRNA than found with control CMV-GFP (Fig. 2, D and E; P < 0.05, n = 3). As observed for A3, haplotypes A1-A4 decreased the steady-state levels of GFP mRNA, compared with cells transfected with control CMV-GFP, both, in resting (Fig. 2, F and G, DMEM; P < 0.01, n = 5) and stimulated cells (Fig. 2G, CD3/CD28; P < 0.01, n = 4). Haplotype A5 did not cause significant differences in the expression of GFP mRNA. These results parallel those analyzing GFP protein (Fig. 2A), and suggest that the mechanism explaining GFP differences between transfectants is at the level of mRNA half-life.
Haplotypes A1-A4 increase the decay rate of GFP mRNA
The above results suggest that A5 and A1-A4 haplotypes differ in their capacity to posttranscriptionally regulate gene expression by a mechanism, depending on the stability of GFP mRNA. To further clarify whether the mRNA decay rate is enhanced in A1-A4 transfectants, we examined the rate of GFP-TNFR2 mRNA decay after inhibition of RNA polymerase II-dependent transcription by ActD (47). Jurkat cells were transfected with constructs containing A1, A2, A3, A4, A5, and CMV-GFP. Twenty-four hours later, cells were treated with ActD to arrest transcription. Cells were harvested at distinct time points over a 3-h period, and GFP mRNA levels were analyzed by RPA. The results show that the mRNAs of GFP-A1, GFP-A2, GFP-A3, and GFP-A4 analyzed together (A1-A4, Fig. 3, A and B), although relatively stable (half-life 4.0 h; P < 0.01, n = 4), exhibit a faster decay rate than the control CMV-GFP directed by the ?-globin UTR. Therefore, the differences in GFP expression between cells transfected with GFP-A1 to GFP-A4 and cells transfected with control CMV-GFP (Fig. 2A) can be explained in terms of mRNA decay rates. The decay curve of GFP-A5 mRNA (half-life 5.8 h) showed a trend toward increased decay in comparison with control CMV-GFP, although the differences were not statistically significant.
FIG. 3. ActD time-course studies in Jurkat T cells. A and C, Analysis of GFP mRNA decay by RPA. Jurkat cells were transfected with GFP constructs and treated with 8 μg/ml ActD (A) or aCD3/aCD28 for 3 h followed by 8 μg/ml ActD (C). Cells were harvested at the times indicated after ActD treatment and lysed. RPA analysis was performed directly in guanidium thiocyanate lysates. Digested RNA was resolved on a 6% nondenaturing polyacrylamide gel, which was then vacuum dried and expression of GFP mRNA and cellular GAPDH analyzed by phosphor imaging using a Bio-Rad molecular imager. A1-A4, Cells transfected with individual plasmids containing haplotypes A1, A2, A3, or A4 mixed and treated with ActD. CMV-GFP. A5, Cells from four independent transfections with plasmids CMV-GFP or GFP-A5, respectively, were mixed and treated with ActD. Arrows indicate the position of the protected RNAs. B and D, GFP mRNA half-life. The expression of GFP mRNA and cellular GAPDH were determined by phosphor imaging of bands in A and C and the GFP to GAPDH ratio represented as a percentage of the zero-time value for each series. The slope of A1-A4 differed from that of CMV-GFP (B: **, P < 0.01, n = 4). The slope of both A1-A4 and A5 differed from CMV-GFP (D: *, P < 0.05; **, P < 0.01, n = 4). E, Binding of Jurkat cytosolic protein(s) to the 3'-UTR of TNFR2. The ability of the TNFR2 3'-UTR to be recognized by Jurkat cytosolic proteins was tested by RNA EMSA. Radiolabeled TNFR2 A2 and A5 RNAs were incubated with Jurkat cell extract and digested with RNase T1 (150 U). Complexes were resolved on a nondenaturing polyacrylamide gel. Arrows indicate RNA-protein complexes. Arrowhead indicates digested RNA. Asterisk indicates free undigested RNA. Data are representative of results from three independent experiments.
Raghavan et al. (48) recently reported that TNFR2 mRNA is destabilized in T lymphocytes on stimulation through the T-cell receptor. The results of Fig. 3B suggest that the TNFR2 3'-UTR fragment studied here contains cis-element(s) with destabilizing capacity in unstimulated Jurkat T cells. To determine whether this destabilizing capacity is regulated by stimulation, half of the cells transfected with the various constructs were incubated with a combination of aCD3 and aCD28 antibodies to fully activate the T cells (49, 50, 51). After 3 h of stimulation, transcription was stopped with ActD, and cells were harvested at the same time points as in the earlier experiments with this reagent. GFP mRNA levels were determined by RPA (Fig. 3, C and D). Results show a faster decay pattern than in unstimulated cells; this confirms the results of previous studies (48) and suggests that the TNFR2 mRNA fragment studied here contains cis-elements that regulate mRNA decay. The mixture of cells individually transfected with A1-A4 showed a GFP mRNA decay rate of approximately 2.2 h (P < 0.01, compared with CMV-GFP, n = 4), slightly faster than that of GFP-A5 mRNA (3.1 h; P < 0.05, compared with CMV-GFP, n = 4). As expected, the decay rate of control CMV-GFP mRNA did not increase on stimulation of Jurkat cells. These results show that the decay rate of GFP mRNA due to the effect of haplotypes A1-A5 can be modulated by stimulation of Jurkat T cells.
The expression of GFP and its mRNA by transfected cells stimulated with aCD3/aCD28 was higher than in untreated cells (Fig. 4). However, a similar increase in GFP expression was common to all constructs, including control CMV-GFP (Fig. 4A). This was probably caused by a transcriptional effect of stimulation on the CMV promoter. In fact, the augmented expression of CMV-GFP caused by treatment with aCD3/aCD28 was at the mRNA level, without any change in its mRNA decay rate, indicating an increase in CMV promoter activity (Fig. 4, C and D; see also Fig. 3, A and C).
Protein binding properties of the TNFR2 3'-UTR
The results described above suggest the presence of cis-element(s) in the region of the TNFR2 3'-UTR fragment studied here that are able to increase reporter mRNA decay. To assess whether the TNFR2 3'-UTR is specifically recognized by cytosolic proteins, radiolabeled RNA probes corresponding to haplotypes A2 and A5 were generated, incubated with Jurkat cytosolic extracts, and analyzed by gel retardation assay. RNAs containing the A2 and A5 sequences were bound by Jurkat cytosolic proteins, generating protein-RNA complexes of similar mobility (Fig. 3E).
Haplotype A2 does not alter the transcriptional activity of the CMV promoter
The results described above indicate that the 3'-UTR of TNFR2 contains cis-element(s) able to augment reporter mRNA decay. To further evaluate whether there is a transcriptional contribution to the observed effect of haplotypes A1-A4, we performed nuclear run-on assays in Jurkat cells. Labeled transcripts were prepared from cells transfected with GFP-A2 or CMV-GFP and were hybridized to the probes shown in Fig. 4E. The expression of the GFP gene was similar in cells transfected with GFP-A2 or CMV-GFP (Fig. 4E), indicating similar transcription rates. Because the endogenous TNFR2 gene is silent in unstimulated Jurkat T cells (10, 52) (data not shown), a probe containing the TNFR2 3'-UTR fragment was used to determine the specificity of the assay. Of the two transcripts, GFP-A2 and CMV-GFP, only the former contains TNFR2 sequences. Accordingly, the signal detected with the TNFR2 probe in nuclear RNA prepared from cells transfected with CMV-GFP determines the background level of the assay (Fig. 4E).
These results suggest that the decrease in steady-state mRNA levels observed for GFP-A1, GFP-A2, GFP-A3, and GFP-A4 is due to posttranscriptional regulation, resulting in an enhanced rate of decay.
Discussion
Adenylate/uridylate-rich elements (AREs) are often found in the 3'-UTRs of posttranscriptionally regulatedmRNAs. Each ARE may represent a combination of structurally distinct domains, such as AU-rich motifs, U-rich elements, and poly(C) repeats, and the combination of these sequence elements determines its ultimate destabilizing function (53, 54, 55, 56, 57, 58). The TNFR2 3'-UTR contains no AU-rich motif. Instead, like the c-jun ARE, the representative of class III AREs, it contains numerous U/purine dinucleotides. GUUUG repeats like those present in TNFR2 3'-UTR also affect the stability of mRNAs (59, 60, 61, 62, 63, 64, 65). The GUUUG repeats in TNFR2 are close to the polymorphic region and are good candidates for functional cis-elements that control decay. Therefore, in this study a TNFR2 3'-UTR segment containing both the polymorphic region and GUUUG repeats was fused downstream of the GFP gene, and its influence on the stability of the chimeric transcript was assessed.
We have shown that the TNFR2 3'-UTR decreases GFP expression. Haplotypes A1, A2, A3, and A4 lowered GFP to a similar extent, more strongly than A5 (Fig. 2A). The mechanism used by A1-A4 to decrease reporter expression, compared with A5, could be a reduction in translation and/or stability of the GFP mRNA. The analysis of mRNA steady-state (Fig. 2G) points to a regulation of mRNA stability as its predominant target because significant differences were observed between transfectants A1-A4 and control CMV-GFP but not between A5 and CMV-GFP. The hypothesis that the mRNA decay rate is enhanced in A1-A4 transfectants is further sustained by the analysis of mRNA stability. Because the differences studied are small (1.5 times, Fig. 2G), it was not expected to observe extensive variation in the mRNA stability curve. In fact, a trend toward increased mRNA decay rates was found in A1-A4 (half-life 4.0 h), compared with A5 (half-life 5.8 h) (Fig. 3B). These data strengthened our previous conclusion that the A1-A4 haplotypes do decrease GFP expression more than A5, and their effect is most evident at the level of regulation of GFP mRNA steady-state.
TNFR2 transcript is stable (half-life > 27 h) in resting T cells but is destabilized in a stimulus-dependent manner (48). Consistent with the results of Raghavan et al. (48), we have shown that the instability conferred by 3'-UTR sequences increases on T-cell activation (Fig. 4). In both cases, cells transfected with A1-A4 or A5, showed increased mRNA decay rates, compared with control CMV-GFP (Fig. 3D). However, a trend toward higher stability of mRNA conferred by A5 is observed.
The decrease in steady-state GFP mRNA levels caused by the A1-A4 haplotypes and the fact that A5 does show a similar effect in stimulated cells suggests that a common determinant of instability is present in the TNFR2 3'-UTR. This could correspond to the U-rich cluster. In fact, the binding studies performed with the 3'-UTR fragment and Jurkat cell extracts demonstrate the existence of cellular factors that recognize the 3'-UTR of TNFR2, without appreciable differences between A5 and non-A5 alleles (Fig. 3E). Scanning this region to search for additional putative functional elements (66) revealed a CU-rich pattern related to 15-lipoxygenase-differentiation control element (67). The repeated CU-rich sequences were apposed in the secondary structure predicted for the 3'-UTR fragment (Fig. 5). These sequences could be recognized by K homology domain-containing RNA-binding proteins, which are able to control mRNA stability (68). The polymorphism 1690 T/C is close to the CU-rich sequences and could modulate the binding of factors to the CU-elements. However, both A2 and A5 contain a cytosine in position 1690 but conduce reporter expression to a different extent, suggesting that the contribution of nucleotide 1690 alone is not sufficient to determine the observed functional differences. Therefore, it is assumed that at least two polymorphisms act together in modulating trans-acting factor binding or accessibility to the U-rich cluster. Experiments aimed at supershifting the observed RNA-protein complexes with specific antibodies should help to characterize the proteins interacting with the TNFR2 3'UTR.
FIG. 5. Secondary structure of the TNFR2 3'-UTR 485-nt fragment. The predicted secondary structure of the TNFR2 RNA haplotype A5 is shown. The position of the three polymorphisms are circled dark. The location of CU-repeats is indicated by lines.
The phylogenetic conservation of the TNFR2 mRNA 3'-UTR is poor. Both mouse and human TNFR2 mRNAs contain long 3'-UTRs of more than 1000 nt. However, the region corresponding to the polymorphisms in the proximal 3'-UTR region is not well conserved, suggesting that this region is unlikely to contain prevalent cis-acting elements.
Polymorphisms in the 3'-UTR of the TNFR2 gene may contribute to the increased risk of Crohn’s disease (28), multiple sclerosis (69), and insulin resistance and obesity (27). Our results suggest that carriers of the common haplotypes A1-A4 will not show major differences in expression of cell membrane-associated TNFR2. However, A5 carriers may express increased levels of TNFR2.
Taken together, our results suggest that dysfunction of mRNA destabilizing signals that affect steady-state levels of TNFR2 mRNA could modify the abundance of TNFR2 mRNA in adipose tissue. We propose that an increased rate of TNFR2 mRNA decay protects cells from unrestrained TNF effects and that this protection will be diminished in A5 subjects (18). In a previous study, we found that younger subjects with the A5-allele (5.2% of nondiabetic subjects) show significantly higher body mass index (BMI) and leptin levels than subjects who do not carry this allele (see Ref. 27 , reported within A2 subjects). Of interest was the relationship between sTNFR2 levels and BMI and between sTNFR2 and leptin in those subjects in which the TNFR2 gene variants determined differences in BMI and leptin. Together with the present findings, we can speculate that increased BMI and leptin in A5 subjects is due to increased TNF effects. Consistent with this, levels of RNA transcripts encoding TNFR2 are significantly increased in the adipose tissues of obese diabetic animals. In muscle from these diabetic animals, RNA transcripts encoding both TNFR1 and TNFR2 were also significantly elevated, although the increase in TNFR2 transcript abundance was less marked than in fat. Overexpression of mRNA for TNF and both of its receptors could be at least partly normalized by treatment of the diabetic animals with the insulin-sensitizing agent pioglitazone (20).
Molecular characterization of proteins recognizing TNFR2 3'-UTR will help to determine the mechanism through which haplotypes differently regulate reporter mRNA decay. Furthermore, transfection analysis of mutants with deletions of the U-rich and C-rich elements will contribute to elucidating the activity of this putative ARE. We are currently performing additional studies to examine the functionality of these polymorphisms.
Acknowledgments
The authors are grateful to the flow cytometry team of the Serveis CientificoTècnics (Universitat de Barcelona) for help in these studies.
References
Vasalli P 1992 The pathophysiology of tumor necrosis factors. Annu Rev Immunol 10:411–452
Vandenabeele P, Declercq W, Beyaert R, Fiers W 1995 Two tumour necrosis factor receptors: structure and function. Trends Cell Biol 5:392–399
Ware CF, Crowe PD, Vanarsdale TL, Andrews JL, Grayson MH, Jerzy R, Smith CA, Goodwin RG 1991 Tumor necrosis factor (TNF) receptor expression in T lymphocytes. Differential regulation of the type I TNF receptor during activation of resting and effector T cells. J Immunol 147:4229–4238
Gehr G, Gentz R, Brockhaus M, Loetscher H, Lesslauer W 1992 Both tumor necrosis factor receptor types mediate proliferative signals in human mononuclear cell activation. J Immunol 149:911–917
Tartaglia LA, Goeddel DV, Reynolds C, Figari IS, Weber RF, Fendly BM, Palladino Jr MA 1993 Stimulation of human T-cell proliferation by specific activation of the 75-kDa tumor necrosis factor receptor. J Immunol 151:4637–4641
Zheng L, Fisher G, Miller RE, Peschon J, Lynch DH, Lenardo MJ 1995 Induction of apoptosis in mature T cells by tumour necrosis factor. Nature 377:348–351
Grell M, Becke FM, Wajant H, Mannel DN, Scheurich P 1998 TNF receptor type 2 mediates thymocyte proliferation independently of TNF receptor type 1. Eur J Immunol 28:257–263
Alexander-Miller MA, Derby MA, Sarin A, Henkart PA, Berzofsky JA 1998 Supraoptimal peptide-major histocompatibility complex causes a decrease in bc1–2 levels and allows tumor necrosis factor receptor II-mediated apoptosis of cytotoxic T lymphocytes. J Exp Med 188:1391–1399
Herbein G, Mahlknecht U, Batliwalla F, Gregersen P, Pappas T, Butler J, O’Brien WA, Verdin E 1998 Apoptosis of CD8+ T cells is mediated by macrophages through interaction of HIV gp120 with chemokine receptor CXCR4. Nature 395:189–194
Pimentel-Muinos FX, Seed B 1999 Regulated commitment of TNF receptor signaling: a molecular switch for death or activation. Immunity 11:783–793
Seckinger P, Isaaz S, Dayer JM 1989 Purification and biologic characterization of a specific tumor necrosis factor inhibitor. J Biol Chem 264:11966–11973
Engelmann H, Novick D, Wallach D 1990 Two tumor necrosis factor-binding proteins purified from human urine. Evidence for immunological cross-reactivity with cell surface tumor necrosis factor receptors. J Biol Chem 265:1531–1536
Schroder J, Stuber F, Gallati H, Schade FU, Kremer B 1995 Pattern of soluble TNF receptors I and II in sepsis. Infection 23:143–148
Gabay C, Cakir N, Moral F, Roux-Lombard P, Meyer O, Dayer JM, Vischer T, Guerne PA 1997 Circulating levels of tumor necrosis factor soluble receptors in systemic lupus erythematosus are significantly higher than in other rheumatic diseases and correlate with disease activity. J Rheumatol 24:303–308
Cope AP, Aderka D, Doherty M, Engelmann H, Gibbons D, Jones AC, Brennan FM, Maini RN, Wallach D, Feldmann M 1992 Increased levels of soluble tumor necrosis factor receptors in the sera and synovial fluid of patients with rheumatic diseases. Arthritis Rheum 35:1160–1169
Douni E, Kollias G 1998 A critical role of the p75 tumor necrosis factor receptor (p75TNF-R) in organ inflammation independent of TNF, lymphotoxin , or the p55TNF-R. J Exp Med 188:1343–1352
Hotamisligil GS 1999 Mechanisms of TNF--induced insulin resistance. Exp Clin Endocrinol Diabetes 107:119–125
Fernández-Real JM, Ricart W 2003 Insulin resistance and chronic cardiovascular inflammatory syndrome. Endocr Rev 24:278–301
Uysal KT, Wiesbrock SM, Hotamisligil GS 1998 Functional analysis of tumor necrosis factor (TNF) receptors in TNF--mediated insulin resistance in genetic obesity. Endocrinology 139:4832–4838
Hofmann C, Lorenz K, Braithwaite SS, Colca JR, Palazuk BJ, Hotamisligil GS, Spiegelman BM 1994 Altered gene expression for tumor necrosis factor- and its receptors during drug and dietary modulation of insulin resistance. Endocrinology 134:264–270
Kern PA, Saghizadeh M, Ong JM, Bosch RJ, Deem R, Simsolo RB 1995 The expression of tumor necrosis factor in human adipose tissue. Regulation by obesity, weight loss, and relationship to lipoprotein lipase. J Clin Invest 95:2111–2119
Hotamisligil GS, Arner P, Caro JF, Atkinson RL, Spiegelman BM 1995 Increased adipose tissue expression of tumor necrosis factor- in human obesity and insulin resistance. J Clin Invest 95:2409–2415
Saghizadeh M, Ong JM, Garvey WT, Henry RR, Kern PA 1996 The expression of TNF by human muscle. Relationship to insulin resistance. J Clin Invest 97:1111–1116
Hotamisligil GS, Arner P, Atkinson RL, Spiegelman BM 1997 Differential regulation of the p80 tumor necrosis factor receptor in human obesity and insulin resistance. Diabetes 46:451–455
Fernández-Real JM, Broch M, Ricart W, Casamitjana R, Gutierrez C, Vendrell J, Richart C 1998 Plasma levels of the soluble fraction of tumor necrosis factor receptor 2 and insulin resistance. Diabetes 47:1757–1762
Pantelidis P, Lympany PA, Foley PJ, Fanning GC, Welsh KI, du Bois RM 1999 Polymorphic analysis of the high-affinity tumor necrosis factor receptor 2. Tissue Antigens 54:585–591
Fernandez-Real JM, Vendrell J, Ricart W, Broch M, Gutierrez C, Casamitjana R, Oriola J, Richart C 2000 Polymorphism of the tumor necrosis factor- receptor 2 gene is associated with obesity, leptin levels, and insulin resistance in young subjects and diet-treated type 2 diabetic patients. Diabetes Care 23:831–837
Sashio H, Tamura K, Ito R, Yamamoto Y, Bamba H, Kosaka T, Fukui S, Sawada K, Fukuda Y, Tamura K, Satomi M, Shimoyama T, Furuyama J-I 2002 Polymorphisms of the TNF gene and the TNF receptor superfamily member 1B gene are associated with susceptibility to ulcerative colitis and Crohn’s disease, respectively. Immunogenetics 53:1020–1027
Glenn CL, Wang WY, Benjafield AV, Morris BJ 2000 Linkage and association of tumor necrosis factor receptor 2 locus with hypertension, hypercholesterolemia and plasma shed receptor. Hum Mol Genet 12:1943–1949
Geurts JM, Janssen RG, van Greevenbroek MM, van der Kallen CJ, Cantor RM, Bu X, Aouizerat BE, Allayee H, Rotter JI, de Bruin TW 2000 Identification of TNFRSF1B as a novel modifier gene in familial combined hyperlipidemia. Hum Mol Genet 9:2067–2074
Benjafield AV, Wang XL, Morris BJ 2001 Tumor necrosis factor receptor 2 gene (TNFRSF1B) in genetic basis of coronary artery disease. J Mol Med 79:109–115
Spotila LD, Rodriguez H, Koch M, Adams K, Caminis J, Tenenhouse HS, Tenenhouse A 2000 Association of a polymorphism in the TNFR2 gene with low bone mineral density. J Bone Miner Res 15:1376–1383
Peral B, San Millan JL, Castello R, Moghetti P, Escobar-Morreale HF 2002 The methionine 196 arginine polymorphism in exon 6 of the TNF receptor 2 gene (TNFRSF1B) is associated with the polycystic ovary syndrome and hyperandrogenism. J Clin Endocrinol Metab 87:3977–3983
van der Velden AW, Voorma HO, Thomas AA 2001 Vector design for optimal protein expression. Biotechniques 31:572, 574, 576–580
Waskiewicz AJ, Johnson JC, Penn B, Mahalingam M, Kimball SR, Cooper JA 1999 Phosphorylation of the cap-binding protein eukaryotic translation initiation factor 4E by protein kinase Mnk1 in vivo. Mol Cell Biol 19:1871–1880
Kar-Roy A, Dong W, Michael N, Li Y 2000 Green fluorescence protein as a transcriptional reporter for the long terminal repeats of the human immunodeficiency virus type 1. J Virol Methods 84:127–138
Espel E, García-Sanz JA, Aubert V, Menoud V, Sperisen P, Fernández N, Spertini F 1996 Transcriptional and translational control of TNF- gene expression in human monocytes by major histocompatibility complex class II ligands. Eur J Immunol 26:2417–2424
Kaabache T, Barraud B, Feldmann G, Bernuau D, Lardeux B 1995 Direct solution hybridization of guanidine thiocyanate-solubilized cells for quantitation of mRNAs in hepatocytes. Anal Biochem 232:225–230
Konarska MM, Sharp PA 1986 Electrophoretic separation of complexes involved in the splicing of precursors to mRNAs. Cell 46:845–855
Leibold EA, Munro HN 1988 Cytoplasmic protein binds in vitro to a highly conserved sequence in the 5' untranslated region of ferritin heavy- and light-subunit mRNAs. Proc Natl Acad Sci USA 85:2171–2175
Sambrook J, Russell DW 2001 Molecular cloning. In: Argentine J, ed. Molecular cloning: a laboratory manual. 3rd ed. New York: Cold Spring Harbor Laboratory Press; 17.23–17.29
Mathews DH, Sabina J, Zuker M, Turner DH 1999 Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J Mol Biol 288:911–940
Zuker M 2003 Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res 31:3406–3415
Forsberg EC, Zaboikina TN, Versaw WK, Ahn NG, Bresnick EH 1999 Enhancement of ?-globin locus control region-mediated transactivation by mitogen-activated protein kinases through stochastic and graded mechanisms. Mol Cell Biol 19:5565–5575
Magness ST, Tugores A, Brenner DA 2000 Analysis of ferrochelatase expression during hematopoietic development of embryonic stem cells. Blood 95:3568–3577
Raineri I, Wegmueller D, Gross B, Certa U, Moroni C 2004 Roles of AUF1 isoforms, HuR and BRF1 in ARE-dependent mRNA turnover studied by RNA interference. Nucleic Acids Res 32:1279–1288
Harrold S, Genovese C, Kobrin B, Morrison SL, Milcarek C 1991 A comparison of apparent mRNA half-life using kinetic labeling techniques vs decay following administration of transcriptional inhibitors. Anal Biochem 198:19–29
Raghavan A, Ogilvie RL, Reilly C, Abelson ML, Raghavan S, Vasdewani J, Krathwohl M, Bohjanen PR 2002 Genome-wide analysis of mRNA decay in resting and activated primary human T lymphocytes. Nucleic Acids Res 30:5529–5538
Mittrücker HW, Shahinian A, Bouchard D, Kündig TM, Mak TW 1996 Induction of unresponsiveness and impaired T cell expansion by staphylococcal enterotoxin B in CD28-deficient mice. J Exp Med 183:2481–2488
Saha B, Harlan DM, Lee KP, June CH, Abe R 1996 Protection against lethal toxic shock by targeted disruption of the CD28 gene. J Exp Med 183:2675–2680
Bachmann MF, McKall-Faienza K, Schmits R, Bouchard D, Beach J, Speiser DE, Mak TW, Ohashi PS 1997 Distinct roles for LFA-1 and CD28 during activation of naive T cells: adhesion versus costimulation. Immunity 7:549–557
Chan FK, Lenardo MJ 2000 A crucial role for p80 TNF-R2 in amplifying p60 TNF-R1 apoptosis signals in T lymphocytes. Eur J Immunol 30:652–660
Lagnado CA, Brown CY, Goodall GJ 1994 AUUUA is not sufficient to promote poly(A) shortening and degradation of an mRNA: the functional sequence within AU-rich elements may be UUAUUUA(U/A)(U/A). Mol Cell Biol 14:7984–7995
Xu N, Chen C-YA, Shyu A-B 1997 Modulation of the fate of cytoplasmic mRNA by AU-rich elements: key sequence features controlling mRNA deadenylation and decay. Mol Cell Biol 17:4611–4621
Czyzyk-Krzeska MF, Bendixen AC 1999 Identification of the poly(C) binding protein in the complex associated with the 3' untranslated region of erythropoietin messenger RNA. Blood 93:2111–2120
Paulding WR, Czyzyk-Krzeska MF 1999 Regulation of tyrosine hydroxylase mRNA stability by protein-binding, pyrimidine-rich sequence in the 3'-untranslated region. J Biol Chem 274:2532–2538
Tillmann-Bogush M, Heaton JH, Gelehrter TD 1999 Cyclic nucleotide regulation of PAI-1 mRNA stability. Identification of cytosolic proteins that interact with an A-rich sequence. J Biol Chem 274:1172–1179
Shao R, Shi Z, Gotwals PJ, Koteliansky VE, George J, Rockey DC 2003 Cell and molecular regulation of endothelin-1 production during hepatic wound healing. Mol Biol Cell 14:2327–2341
Shimizu N, Ohta M, Fujiwara C, Sagara J, Mochizuki N, Oda T, Utiyama H 1991 Expression of a novel immediate early gene during 12-O-tetradecanoylphorbol-13-acetate-induced macrophagic differentiation of HL-60 cells. J Biol Chem 266:12157–12161
You Y, Chen CY, Shyu AB 1992 U-rich sequence-binding proteins (URBPs) interacting with a 20-nucleotide U-rich sequence in the 3' untranslated region of c-fos mRNA may be involved in the first step of c-fos mRNA degradation. Mol Cell Biol 12:2931–2940
Levine TD, Gao F, King PH, Andrews LG, Keene JD 1993 Hel-N1: an autoimmune RNA-binding protein with specificity for 3' uridylate-rich untranslated regions of growth factor mRNAs. Mol Cell Biol 13:3494–3504
Chen CYA, Shyu AB 1994 Selective degradation of early-response-gene mRNAs: functional analyses of sequence features of the AU-rich elements. Mol Cell Biol 14:8471–8482
Peng H, Lever JE 1995 Regulation of Na(+)-coupled glucose transport in LLC-PK1 cells. Message stabilization induced by cyclic AMP elevation is accompanied by binding of a M(r) = 48,000 protein to a uridine-rich domain in the 3'-untranslated region. J Biol Chem 270:23996–24003
Peng SS, Chen CY, Shyu AB 1996 Functional characterization of a non-AUUUA AU-rich element from the c-jun proto-oncogene mRNA: evidence for a novel class of AU-rich elements. Mol Cell Biol 16:1490–1499
Paillard L, Legagneux V, Maniey D, Osborne HB 2002 c-Jun ARE targets mRNA deadenylation by an EDEN-BP (embryo deadenylation element-binding protein)-dependent pathway. J Biol Chem 277:3232–3235
Pesole G, Liuni S 1999 Internet resources for the functional analysis of 5' and 3' untranslated regions of eukaryotic mRNAs. Trends Genet 15:378
Ostareck-Lederer A, Ostareck DH, Standart N, Thiele BJ 1994 Translation of 15-lipoxygenase mRNA is inhibited by a protein that binds to a repeated sequence in the 3' untranslated region. EMBO J 13:1476–1481
Kiledjian M, Wang X, Liebhaber SA 1995 Identification of two KH domain proteins in the -globin mRNP stability complex. EMBO J 14:4357–4364
Ehling R, Gassner C, Lutterotti A, Strasser-Fuchs S, Kollegger H, Kristoferitsch W, Reindl M, Berger T 2004 Genetic variants in the tumor necrosis factor receptor II gene in patients with multiple sclerosis. Tissue Antigens 63:28–33(Irene Puga, Bego?a Lainez)
Address all correspondence and requests for reprints to: Dr. Enric Espel, Department Fisiologia, Facultat Biologia, Diagonal 645, 08028 Barcelona, Spain. E-mail: eespel@ub.edu.
Abstract
The gene encoding the human TNF receptor (TNFR) 2 contains polymorphisms in the 3' untranslated region (UTR). Previous studies have shown that some variant alleles in this region are associated with obesity and insulin resistance. However, the effect of these polymorphisms on the expression of TNFR2 has not been studied to date. To examine the role played by different haplotypes in the control of TNFR2 expression (haplotypes A1-A5, referring to nucleotides 1663 G/A, 1668 T/G, and 1690 T/C), we introduced these sequences into the 3'-UTR of a heterologous reporter gene and expressed the corresponding constructs in a human T-cell line. We demonstrate that a 485-nt fragment of the TNFR2 3'-UTR that contains a U-rich region decreases reporter expression and that haplotypes A1-A4 exert a stronger effect than A5. Furthermore, time-course assays of mRNA stability using actinomycin D revealed that haplotypes A1-A4 destabilize the mRNA. The proximal TNFR2 3'-UTR, independently of haplotype differences, responded to T-cell activation by increasing mRNA decay. Electromobility shift analysis demonstrated that protein(s) found in T-cell extracts bind to the 485-nt fragment. We suggest that an increased rate of TNFR2 mRNA decay protects cells from unrestrained TNF effects and that this protection is weakened in A5 subjects. These findings may explain the association of this haplotype with obesity and increased leptin levels.
Introduction
TNF IS A CYTOKINE THAT plays a central role in inflammatory phenomena and is involved in the pathogenesis of chronic local or systemic inflammation in vivo (1). The effects of TNF are signaled by two cell surface receptors, TNFR1 (TNFRSF1A or p55) and TNFR2 (TNFRSF1B or p75/80), which mediate cellular responses ranging from apoptosis to cell proliferation and differentiation (2). TNFR2 is predominantly expressed in cells of the hematopoietic lineage and is the principal TNFR expressed by blood mononuclear cells and activated T cells (2, 3, 4), playing a key role in T-cell homeostasis (5, 6, 7, 8, 9, 10).
These two receptors are also produced in a soluble form. This form constitutes the extracellular portion of membrane-associated TNFRs, shed by proteolytic cleavage (11, 12). Soluble TNFR2 (sTNFR2) inhibits TNF action by competing with cell surface receptors to bind TNF. Chronic enhanced production of sTNFR2 is associated with several human inflammatory and autoimmune conditions, including sepsis (13), lupus (14), and rheumatoid arthritis (15). A ligand-independent role for this receptor has also been reported. Thus, transgenic mice that constitutively express enhanced levels of a wild-type human TNFR2 suffer from a severe multiorgan inflammatory syndrome (16).
The role of the TNF-TNFR pathway in insulin resistance and lipid abnormalities has been extensively described (17, 18). Mice with targeted null mutations in the genes that encode TNFR1 and 2 show a significant increase in insulin sensitivity (19), which indicates a primary role for the TNF-TNFR pathway in the development of insulin resistance. Levels of RNA transcripts encoding TNFR2 receptors are significantly increased in the fat and muscle of obese diabetic animals (20). The TNF-TNFR system may also contribute to the regulation of insulin action in human obesity (18). In obese humans and insulin-resistant and diabetic subjects, production of TNF in muscle and adipose tissue is higher than in lean individuals (21, 22, 23), a condition that can be reversed by weight loss (21, 22). Furthermore, plasma levels of sTNFR2 are correlated with basal and postload glucose levels and insulin resistance (24, 25).
The 3'-untranslated region (UTR) of human TNFR2 shows a degree of polymorphism (26). An increase in the frequency of some alleles has been correlated with insulin resistance (27) and Crohn’s disease (28). Interestingly, other mutations in this locus have been associated with several components of insulin resistance syndrome, such as hypertension (29), dyslipidemia (30), and cardiovascular disease (31), and with degenerative diseases such as osteoporosis (32). This locus has also been linked to polycystic ovary syndrome, a condition with known defects in insulin activity (33).
To date, no function has been ascribed to the polymorphisms present in the 3'-UTR of TNFR2. To determine whether these polymorphisms affect gene expression, we fused a genomic DNA fragment containing the polymorphic region of haplotypes A1-A5 to a green fluorescent protein (GFP) reporter gene. We then used flow cytometry to study the expression of this gene in live Jurkat T cells. T cells represent an interesting model with which to study the control of TNFR2 expression because these cells are known to regulate TNFR2 levels on activation (3, 6). This means that the biochemical machinery necessary to regulate TNFR2 expression is present in these cells. Therefore, we decided to undertake the present study in lymphoid Jurkat T cells. Our results indicate that haplotypes A1-A4 decrease the expression of TNFR2 (p75/80) by increasing the decay rate of its mRNA.
Materials and Methods
Chemicals and antibodies
DMEM, penicillin, streptomycin, and fetal calf serum (FCS) were purchased from Sigma-Aldrich (Madrid, Spain). FCS was decomplemented at 56 C for 30 min. Actinomycin D (ActD) was purchased from Tocris (Biogen, Madrid, Spain). Anti-GFP was obtained from Roche Diagnostics (Barcelona, Spain), and antihuman actin was from Sigma-Aldrich. Anti-CD (aCD) 3, a mouse IgG2a monoclonal antibody (33–2A3 workshop II; a kind gift from Dr. R. Vilella, Hospital Clínic, Barcelona, Spain), was used as a hybridoma supernatant at a dilution of 1:1000. This dilution was shown to induce maximal TNF secretion in Jurkat T cells (data not shown). aCD28, a mouse IgM monoclonal antibody (CK243; a kind gift from Dr. Pedro Romero, Ludwig Institute for Cancer Research, Lausanne Branch, Switzerland), was used as a hybridoma supernatant at a dilution of 1:20. This dilution was shown to produce maximal TNF secretion in combination with aCD3 in Jurkat T cells (data not shown).
Reporter plasmid construction
A vector containing an enhanced GFP coding sequence flanked by ?-globin UTRs and directed by the cytomegalovirus (CMV) promoter (pCS3glob-GFP-glob; see below) was used to analyze the putative regulatory sequences of the TNFR2 gene. TNFR2 3'-UTR genomic DNA fragments were introduced into pCS3glob-GFP-glob downstream of the GFP coding region and upstream of the ?-globin 3'-UTR.
To generate a reporter construct that contained haplotypes A3 and A5 in a sense orientation and haplotype A2 in an antisense orientation, genomic DNA was isolated from homozygous donors for each TNFR2 haplotype. A 485-bp fragment of the TNFR2 3'-UTR from A2, A3, and A5 donors was generated by PCR using the following primers: 5'-GGGGAGCACCGAAGAGAA and 5'-AAGGATGCTGGGGTTTTCT. The PCR product was cloned between the GFP gene and the globin 3'-UTR (into the EcoRV site) in the vector pCS3glob-GFP-glob. The construct containing haplotype A3 was used to prepare plasmids containing haplotypes A1, A2, and A4 by site-directed mutagenesis. In this way, we obtained haplotypes A1-A5 with the following variations at positions 1663 A/G, 1668 T/G, and 1690 T/C (Table 1), which spread from the stop codon (TAA, Fig. 1) some 188, 193, and 215 nt downstream, respectively (National Center for Biotechnology Information, NM 001066).
TABLE 1. Characterization of variants in the 3'-UTR (exon 10) of the TNFR2 gene and frequency of these variants in the population
FIG. 1. Linearity of response of the GFP reporter in Jurkat T cells. A, Schematic diagram of the GFP reporter gene, indicating the coding region and the ?-globin 3'-UTR fragment. Constructs are based on the vector pCS3+mT (35 ). The TNFR2 haplotypes A1-A5 (the polymorphic nucleotides are indicated) and A2 in antisense orientation (A2as) were assembled downstream of the GFP coding region, as shown. B, Example of FACS analysis with GFP-A5-transfected cells. Histogram showing the fluorescence of cells transfected with 20 μg of plasmid GFP-A5 or with empty vector pCS3+mT (without GFP; dark area). The relative amount of emitted fluorescence from each cell is shown on the horizontal axis. The data were grouped into four intervals, as indicated by the top bar. From these data, the MFI of the sample was obtained by multiplying the percentage of cells within each interval by its MFI and adding the result of the four intervals (36 ). C, Relative expression of CMV-GFP in Jurkat T cells. Cells (1.0 x 107) were transfected with increasing amounts of the CMV-GFP vector along with 5 μg of a luciferase-expressing vector (pGL2). Twenty-four hours later, GFP expression was analyzed in live cells by FACS. The MFI normalized by the transfection efficiency (luciferase) is shown on the y-axis, and the amount of DNA transfected on the x-axis. MIF was calculated as described above (B). Data represent the average of three independent transfections ± SD. D, The sequence of the TNFR2 3'-UTR fragment (485 bp) for haplotype A2 is shown with the position of nine U-rich stretches containing three or more Us underlined. The U-rich cluster is reminiscent of AREs. The three polymorphic nucleotides are shown in lowercase and circled. The 485-bp fragment was inserted into a unique EcoRV restriction site localized between GFP and the globin 3'-UTR (A).
The constructs pCS3glob-GFP-glob and pCS3glob-GFP-A3glob were fully sequenced (ABI PRISM, PerkinElmer, Barcelona, Spain) before use in transfection experiments. Polymorphism 1663 A/G was also controlled by FauI digestion. The rest of the plasmids were sequenced through the GFP-coding and 3'-UTR region. For simplicity, the plasmids are called CMV-GFP (pCS3glob-GFP-glob), GFP-A1 (pCS3glob-GFP-A1-glob), GFP-A2 (pCS3glob-GFP-A2-glob), GFP-A3 (pCS3glob-GFP-A3-glob), GFP-A4 (pCS3glob-GFP-A4-glob), GFP-A5 (pCS3glob-GFP-A5-glob), and GFP-A2as (A2 in antisense orientation, CS3glob-GFP-A2as-glob).
Plasmid CMV-GFP was generated as follows: plasmid pT7globGFPglob, which contains xenopus ?-globin 5'- and 3'-UTRs flanking the GFP sequence, was a gift from Dr. A. A. M. Thomas (Department of Developmental Biology, University of Utrecht, Utrecht, The Netherlands) (34). The fragment globGFPglob was excised by digestion with HindIII and EcoRI and then subcloned into the vector pCS3+mT downstream of the CMV promoter (35).
Transfections
Jurkat T cells were grown in DMEM containing 10% FCS. Cells were diluted to 0.5 x 106 cells/ml the day before transfection. The next day they were pelleted, washed in PBS, counted, and resuspended at a final concentration of 5 x 107 cells/ml in 0.2-cm-wide electroporation cuvettes (Invitrogen, Barcelona, Spain). Transient transfection was performed by electroporation. Briefly, 10 μg (or the amount indicated in figure legends) of GFP constructs and, where indicated, 5 μg of the luciferase reporter plasmid pGL2 (Promega, Innogenetics, Barcelona, Spain) were transfected into cells in a 200 μl cell suspension (1700 μF, 72, 126 V, 30 msec, using a BTX 600). Cells were then seeded into petri dishes at a final density of 5 x 105 cells/ml in DMEM-10% FCS. Experiments were carried out 24 h after transfection. Cells (0.5 x 106 in 1 ml) were taken directly from culture flasks and placed in 3-ml plastic tubes for fluorescence-activated cell sorting (FACS) analysis. One microliter of propidium iodide (0.07%) was then added, and tubes were maintained on ice in the dark before flow cytometry studies. Luciferase activity was measured in cell extracts according to the manufacturer’s instructions (Promega, Innogenetics). The transfection efficiency of most experiments was determined from the percentage of cells displaying green fluorescence (36). This procedure provided more consistent results than cotransfecting luciferase. Transfection efficiency values varied between 5 and 70%.
Flow cytometry
Transgene expression in live cells was measured by flow cytometric analysis on an Epics XL flow cytometer (Coulter Corp., Miami, FL) equipped with a 488-nm argon laser at 15 mW. Analysis was performed according to the method described by Kar-Roy et al. (36). This method offers advantages over traditional luciferase or ?-galactosidase reporter genes in that GFP transgene expression can be analyzed in individual, live-cell populations. The instrument was set up with the standard configuration: forward scatter, side scatter, and green (525 nm) fluorescence for GFP were collected. Green fluorescence was represented on a logarithmic histogram. The relationship between the fluorescence signal and the number of emitting chromophores per cell is linear, a higher signal indicating higher expression levels of the reporter (36).
Twenty-four hours after transfection, the cells (0.5 ml, containing approximately 2.5 x 105 cells) were analyzed. Gating for live cells was achieved by forward scatter and side scatter. The histogram of the intensity of green fluorescence in the live cells was used to analyze GFP expression. Quantification of 3'-UTR regulatory effects on GFP expression was performed by measuring the median of the fluorescence intensity histogram obtained from the transfected population (36). Differences in GFP expression, measured by the median fluorescence intensity (MFI), indicate distinct posttranscriptional regulatory activity of the UTRs tested. Because this system was based on a transient transfection assay, the population of transfected cells was heterogeneous in terms of their GFP expression levels (36). Optical alignment was based on optimized signal from 10 nm fluorescent beads (Flowcheck, Epics Division, Coulter Corp.). Time vs. fluorescence was used as a control for the stability of the instrument. GFP expression results are presented as MFI/transfection efficiency.
Western blotting
Cells (1.0 x 106) were lysed in 100 μl Laemmli sample buffer. The samples were heated at 95 C for 30 min and separated by SDS-PAGE. A 15% gel was used for GFP analysis. The proteins were blotted onto polyvinylidene difluoride membranes and detected by immunoblot. GFP was detected with a mixture of mouse monoclonal antibodies (Roche Diagnostics) at a dilution of 3 μg per 8 ml.
After stripping of anti-GFP (Re-blot plus; Chemicon, Pacisa-Giralt, Madrid, Spain), the same membrane was reprobed with an antiactin polyclonal antibody raised in rabbit, diluted at 1:200 (reference A-2066; Sigma-Aldrich).
Ribonuclease (RNase) protection assay and GFP probes
RNase protection assays (RPA) were performed as described previously (37). Digested RNA was resolved on a 6% polyacrylamide-urea denaturing gel. In some experiments, RPA analysis was performed directly in cell lysate, as described elsewhere. In these cases, digested RNA was fractionated on a 6% polyacrylamide/Tris-borate-EDTA nondenaturing gel and vacuum dried onto nylon membranes (Nytran Supercharge; Sleicher & Schuell, Acefesa, Barcelona, Spain). Radioactivity of protected bands was measured by phosphor imager analysis using a FX molecular imager (Bio-Rad Laboratories, Barcelona, Spain).
To prepare the GFP probe, a GFP DNA fragment was cloned into the pSTBlue-1 vector (Novagen, Bionova, Madrid, Spain). The recombinant plasmid was digested with FokI. T7 polymerase generates an antisense RNA of 494 nt from the FokI-digested plasmid. RNase digestion of samples that contain glob-GFP-glob mRNA hybridized to the antisense RNA probe generates a protected probe fragment of 338 nt.
The probe used in the RPA (see Fig. 4) was prepared as follows: a DNA fragment of 254 bp was generated by PCR from the plasmid CMV-GFP using the primers GFP forward AGCAAGGGCGAGGAG and T7-GFP reverse GAATTCTAATACGACTCACTATAGGTGTGGTCGGGGTAGC. T7 polymerase was then used to generate an antisense RNA of 231 nt from that fragment. The protected fragment after hybridization with GFP RNA and digestion with RNases A and T1 is 229 nt in length, a 2-nt difference from the nondigested one.
FIG. 4. Stimulation of Jurkat cells does not reverse the inhibition of GFP expression by the TNFR2 3'-UTR. A, Jurkat cells were transfected with GFP-TNFR2 constructs, as indicated in Fig. 2. Cells were starved by culturing in DMEM-0.1% FCS. Twenty hours later, cells were stimulated with aCD3 and aCD28 antibodies for 5 h, and GFP expression was analyzed by FACS. Results are shown as fold increase over control and nonstimulated cells. Experiments were performed in duplicate; the data shown are the mean values of at least three independent transfections ± SD. B, C, and D, Jurkat cells were transfected with CMV-GFP and stimulated with aCD3/aCD28 antibodies for 1 h, and GFP expression was analyzed by FACS (B) and RPA (C and D). The labeled antisense RNA probe used for GFP in this RPA was 231 nt long, generated as indicated in Materials and Methods. Arrowheads (D) indicate the positions of the undigested probes. Arrows indicate the positions of protected fragments. E, Nuclear run-on analysis of GFP gene activity in transfected cells. Total [32P]RNA isolated from Jurkat T cells transfected with plasmid GFP-A2 or CMV-GFP was hybridized to slot blots containing TNFR2 3'-UTR DNA or GFP DNA, as described in Materials and Methods. Bars, obtained by phosphor imaging of slot blots, represent the relative transcription rates of GFP-A2 and CMV-GFP genes. Data represent the mean of two independent experiments.
GFP mRNA decay
For the GFP mRNA decay experiments, transfected cells were incubated for 12 h at 37 C in DMEM-0.2% FCS. aCD3 and aCD28 monoclonal antibodies were added and cells were incubated at 37 C for a further 3 h. ActD (8 μg/ml) was added to the culture dishes and cells incubated for varying lengths of time, up to 3 h. Cells were pelleted by centrifugation and lysed with 0.4 ml of 4 M guanidium thiocyanate, 25 mM Na-citrate (pH 7.0), 0.5% N-lauryl sarcosine. RPA analysis was performed directly in cell lysates, as described elsewhere (38), using the GFP and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probes generated with T7 RNA polymerase (see RNase protection assay). Radioactivity of protected bands was measured by phosphor imager analysis. The phosphor imager values were represented vs. time with ActD, and the slopes were analyzed by linear regression analysis.
RNA EMSA
Jurkat cell extracts were prepared as described previously (39, 40). Briefly, cells (1.0 x 108) were pelleted, washed with PBS, resuspended in 1 ml PBS, and centrifuged at 13,000 x g for 10 sec in Eppendorf tubes to remove the supernatant. The cell pellet was lysed in 830 μl Munro-lysis buffer [10 mM HEPES (pH 7.6), 3 mM MgCl2, 40 mM KCl, 5% glycerol, 1 mM dithiothreitol, leupeptin, aprotinin] supplemented with 0.2% Nonidet P-40 by pipetting 10 times. The cell lysate was diluted with 1.66 ml of Munro-lysis buffer and centrifuged at 13,000 x g for 2 min to pellet the nuclei. The supernatant (cell extract) was aliquoted and kept frozen at –80 C until used. To prepare RNA probes, the 485-bp TNFR2 3'-UTR fragment was amplified by PCR downstream of the T7 promoter using the oligonucleotides forward GGTAATACGACTCACTATA-GGGGAGCACCGAAGAGAA (T7 promoter linked to TNFR2 3'-UTR) and reverse AAGGATGCTGGGGTTTTCT, using the GFPA1-A5 constructs as templates. The amplified fragments were purified (QiaQuick; QIAGEN, Izasa, Barcelona, Spain) and incubated with T7 RNA polymerase to generate 32P-uridine-labeled RNA probes. Assays consisted of 15 μg of Jurkat cytosolic extract incubated with 50,000 cpm of labeled RNA, 4 μg/μl yeast tRNA, and 2 μg/μl heparin in a total volume of 17 μl. Reactions were incubated at room temperature for 15 min, followed by 10 min with RNase T1 (150 U). Samples were resolved on 4% nondenaturing polyacrylamide gels.
Run-on
To monitor transcriptional activity, nuclear run-on assays were performed as described elsewhere (41). Nonradioactive probes for GFP and TNFR2 were prepared by PCR using the following oligonucleotides: GFP, 5'-AGCAGCACGACTTCTTCA and 5'-CGTTGTGGCTGTTGTATT; TNFR2, 5'-GGGGAGCACCGAAGAGAA and 5'-AAGATGCTGGGGTTTTCT. RNA labeling in run-on transcription mixtures containing 6.0 x 107 nuclei from transfected cells was performed using [-32P]UTP. RNA was extracted with phenol/chloroform/isoamyl alcohol and precipitated with ethanol. Nylon membranes (NYTRAN SuPerCharge; Schleicher & Schüell) containing the nonradioactive probes were hybridized to the labeled RNA at 68 C for 72 h. After hybridization, the membranes were washed once with 2x saline sodium citrate (SSC), 0.1% sodium dodecyl sulfate (SDS) for 5 min at room temperature, twice with 0.5x SSC, 0.1% SDS for 20 min at 68 C, and finally, once with 0.1x SSC, 0.1% SDS for 20 min at 68 C. The radioactivity contained in the bands was determined using a molecular imager (Bio-Rad).
Analysis of RNA secondary structure
RNA secondary structure analysis was performed using the MFOLD program designed by Mathews et al. (42, 43) (available at http://www.bioinfo.rpi.edu/applications/mfold). The folding temperature was fixed at 37 C.
Statistical analysis
Data are expressed as the mean ± SD. Statistically significant differences between samples were determined using a one-factor ANOVA followed by a t test, except for flow cytometry data for which the Bonferroni multiple comparisons test was used.
Results
To evaluate a potential functional role for the polymorphisms located in the 3'-UTR of TNFR2 (Table 1), we introduced a 485-bp sequence that contained the polymorphic region of the TNFR2 3'-UTR into reporter constructs (Fig. 1). TNFR2 sequences were introduced between the GFP gene, used as a quantitative reporter (36, 44, 45, 46), and the stable ?-globin UTR. Because the fluorescence of transiently transfected cells is heterogeneous, the MFI of each sample was calculated by dividing the histogram into four quadrants and determining the fluorescence for each quadrant (Fig. 1B) (36). The fluorescence intensity emitted by cells transfected with constructs containing the GFP gene is proportional to the amount of GFP produced by the transfected cells and increased progressively as a function of the amount of DNA transfected (Fig. 1C), thereby confirming the validity of the procedure.
Haplotypes A1-A4 reduce expression of a GFP reporter
The fluorescence of transfected cells displayed a linear increase with increasing amounts of the different constructs (Fig. 2A). However, insertion of the TNFR2 haplotypes A1, A2, A3, or A4 similarly reduced the expression of GFP to 50% of that observed with CMV-GFP at any of the three DNA concentrations tested (Fig. 2A; P < 0.05, n = 6). The reduction in the expression of GFP by haplotype A5 was less pronounced (35% at 20 μg; P < 0.05, n = 8). Furthermore, expression of GFP-A5 was significantly higher than any of the other haplotypes A1-A4 (observed at 20 μg, P < 0.05, n = 8). As a control, A2 used in an antisense orientation (A2as), displayed similar levels of GFP expression to those found with CMV-GFP.
FIG. 2. Expression of GFP-3'-UTR-TNFR2 in Jurkat T cells. A, Regulation of GFP expression by TNFR2 3'-UTR haplotypes. Cells (1.0 x 107) were transfected with increasing amounts of plasmids containing haplotypes A1-A5. Twenty-four hours later, fluorescence was analyzed in live cells by FACS. The MFI for GFP was calculated and normalized to the transfection efficiency. The normalized ratio (GFP fluorescence to transfection efficiency) is represented as a percent of the maximum value. Measurements were performed in duplicate; the data represent the mean ± SE from at least six independent experiments. *, Significant differences (P < 0.05, n = 6). B, Transfection efficiency. Transfection efficiency in A was determined from the percentage of GFP-positive (+ve) cells in each sample (36 ). The data shown are the mean values of at least six independent experiments ± SD. C, Analysis of GFP expression by Western blot. Cells transfected in A with 20 μg of the indicated constructs were lysed and proteins fractionated by SDS-PAGE (15% gels) and transferred to a membrane. GFP expression was detected with an anti-GFP monoclonal antibody. The same membrane was reprobed with an antiactin antibody. Apparent molecular weight is indicated on the right of the figure. This is a representative experiment from a total of three. D, Analysis of GFP mRNA by RPA. Total RNA was prepared from cells transfected in A and GFP mRNA analyzed by RPA with probes for GFP and cellular GAPDH. Protected RNAs were fractionated on a urea denaturing gel and detected by autoradiography. Arrowheads indicate the position of undigested probes. Arrows indicate the position of protected fragments. Top, Film was overexposed to show differences in GFP expression between samples. P, Probes; Pr, probes treated in the same way as the rest of the samples but without RNases; Pd, probes digested with RNases as for the rest of the samples; –, cells transfected with empty-vector (lacking GFP). The experiment was repeated three times and a representative result is shown. E, GFP mRNA expression in transfected cells. Radioactivity of protected bands in D was measured using a Bio-Rad molecular phosphor imager and the GFP to GAPDH ratio calculated to normalize differences in sample loading. In these experiments, the transfection efficiency was calculated by cotransfection of a luciferase reporter (pGL2). The normalized GFP to GAPDH to luciferase (Luc) ratio is shown as a percent of the maximum value. Data shown are mean values ± SD from three independent experiments. A3 expression differed from that of CMV-GFP (*, P < 0.05). F, Analysis of GFP mRNA by RPA. RPA analysis was performed directly in guanidium thiocyanate lysates of cells transfected in A. Digested RNA was resolved on a 6% nondenaturing polyacrylamide gel, which was then vacuum dried and analyzed by phosphor imaging. Arrows indicate the position of protected fragments. A1-A4, Cells transfected with individual reporter plasmids containing haplotypes A1, A2, A3, or A4 mixed before lysis with guanidium thiocyanate. The experiment was repeated four times, and a representative result is shown. G, GFP mRNA expression in transfected cells. Radioactivity in protected bands from F and similar experiments was measured by phosphor imaging, and the GFP to GAPDH ratio calculated to normalize differences in sample loading. Transfection efficiency was determined from the percentage of GFP +ve cells. The normalized ratio (GFP to GAPDH) to transfection efficiency is shown as a percent of the maximum value. The data are shown as mean values ± SD. A1-A4 expression differed from CMV-GFP, *, Differences statistically significant in nonstimulated (P < 0.01) (DMEM, n = 5) and stimulated conditions (cells stimulated with aCD3+aCD28 antibodies, n = 4).
The amount of the GFP protein in transfected cells was also analyzed by Western blot (Fig. 2C). Gel loading was verified by reprobing with an antiactin antibody. Although loading of cell lysates from samples transfected with GFP constructs containing A1 to A5 or CMV-GFP was similar, the amount of GFP in cells transfected with GFP-A1 to GFP-A4 was lower, confirming the results obtained by analysis of GFP fluorescence.
Haplotypes A1-A4 reduce the expression of GFP mRNA
Because the differences between haplotypes A1-A5 and CMV-GFP reside in the 3'-UTR, the results in Fig. 2A suggest that the observed differences in GFP expression correspond to posttranscriptional mechanisms. This could correspond to differences in mRNA decay and/or mRNA translation. To ascertain which of the two mechanisms was causing the difference in expressed GFP, the level of GFP mRNA was compared between transfectants. Because the four haplotypes A1-A4 behaved similarly in controlling the expression of GFP (Fig. 2, A and C), RNA analysis was performed on cells transfected with GFP-A3 and in a mixture of cells transfected with each of the constructs GFP-A1, GFP-A2, GFP-A3, and GFP-A4. As expected, haplotype A2 in antisense orientation did not modify the GFP mRNA steady-state levels observed when transfecting control CMV-GFP (Fig. 2, D and E). However, the expression of GFP-A3 mRNA followed a pattern similar to that of the GFP protein, with a lower accumulation of this mRNA than found with control CMV-GFP (Fig. 2, D and E; P < 0.05, n = 3). As observed for A3, haplotypes A1-A4 decreased the steady-state levels of GFP mRNA, compared with cells transfected with control CMV-GFP, both, in resting (Fig. 2, F and G, DMEM; P < 0.01, n = 5) and stimulated cells (Fig. 2G, CD3/CD28; P < 0.01, n = 4). Haplotype A5 did not cause significant differences in the expression of GFP mRNA. These results parallel those analyzing GFP protein (Fig. 2A), and suggest that the mechanism explaining GFP differences between transfectants is at the level of mRNA half-life.
Haplotypes A1-A4 increase the decay rate of GFP mRNA
The above results suggest that A5 and A1-A4 haplotypes differ in their capacity to posttranscriptionally regulate gene expression by a mechanism, depending on the stability of GFP mRNA. To further clarify whether the mRNA decay rate is enhanced in A1-A4 transfectants, we examined the rate of GFP-TNFR2 mRNA decay after inhibition of RNA polymerase II-dependent transcription by ActD (47). Jurkat cells were transfected with constructs containing A1, A2, A3, A4, A5, and CMV-GFP. Twenty-four hours later, cells were treated with ActD to arrest transcription. Cells were harvested at distinct time points over a 3-h period, and GFP mRNA levels were analyzed by RPA. The results show that the mRNAs of GFP-A1, GFP-A2, GFP-A3, and GFP-A4 analyzed together (A1-A4, Fig. 3, A and B), although relatively stable (half-life 4.0 h; P < 0.01, n = 4), exhibit a faster decay rate than the control CMV-GFP directed by the ?-globin UTR. Therefore, the differences in GFP expression between cells transfected with GFP-A1 to GFP-A4 and cells transfected with control CMV-GFP (Fig. 2A) can be explained in terms of mRNA decay rates. The decay curve of GFP-A5 mRNA (half-life 5.8 h) showed a trend toward increased decay in comparison with control CMV-GFP, although the differences were not statistically significant.
FIG. 3. ActD time-course studies in Jurkat T cells. A and C, Analysis of GFP mRNA decay by RPA. Jurkat cells were transfected with GFP constructs and treated with 8 μg/ml ActD (A) or aCD3/aCD28 for 3 h followed by 8 μg/ml ActD (C). Cells were harvested at the times indicated after ActD treatment and lysed. RPA analysis was performed directly in guanidium thiocyanate lysates. Digested RNA was resolved on a 6% nondenaturing polyacrylamide gel, which was then vacuum dried and expression of GFP mRNA and cellular GAPDH analyzed by phosphor imaging using a Bio-Rad molecular imager. A1-A4, Cells transfected with individual plasmids containing haplotypes A1, A2, A3, or A4 mixed and treated with ActD. CMV-GFP. A5, Cells from four independent transfections with plasmids CMV-GFP or GFP-A5, respectively, were mixed and treated with ActD. Arrows indicate the position of the protected RNAs. B and D, GFP mRNA half-life. The expression of GFP mRNA and cellular GAPDH were determined by phosphor imaging of bands in A and C and the GFP to GAPDH ratio represented as a percentage of the zero-time value for each series. The slope of A1-A4 differed from that of CMV-GFP (B: **, P < 0.01, n = 4). The slope of both A1-A4 and A5 differed from CMV-GFP (D: *, P < 0.05; **, P < 0.01, n = 4). E, Binding of Jurkat cytosolic protein(s) to the 3'-UTR of TNFR2. The ability of the TNFR2 3'-UTR to be recognized by Jurkat cytosolic proteins was tested by RNA EMSA. Radiolabeled TNFR2 A2 and A5 RNAs were incubated with Jurkat cell extract and digested with RNase T1 (150 U). Complexes were resolved on a nondenaturing polyacrylamide gel. Arrows indicate RNA-protein complexes. Arrowhead indicates digested RNA. Asterisk indicates free undigested RNA. Data are representative of results from three independent experiments.
Raghavan et al. (48) recently reported that TNFR2 mRNA is destabilized in T lymphocytes on stimulation through the T-cell receptor. The results of Fig. 3B suggest that the TNFR2 3'-UTR fragment studied here contains cis-element(s) with destabilizing capacity in unstimulated Jurkat T cells. To determine whether this destabilizing capacity is regulated by stimulation, half of the cells transfected with the various constructs were incubated with a combination of aCD3 and aCD28 antibodies to fully activate the T cells (49, 50, 51). After 3 h of stimulation, transcription was stopped with ActD, and cells were harvested at the same time points as in the earlier experiments with this reagent. GFP mRNA levels were determined by RPA (Fig. 3, C and D). Results show a faster decay pattern than in unstimulated cells; this confirms the results of previous studies (48) and suggests that the TNFR2 mRNA fragment studied here contains cis-elements that regulate mRNA decay. The mixture of cells individually transfected with A1-A4 showed a GFP mRNA decay rate of approximately 2.2 h (P < 0.01, compared with CMV-GFP, n = 4), slightly faster than that of GFP-A5 mRNA (3.1 h; P < 0.05, compared with CMV-GFP, n = 4). As expected, the decay rate of control CMV-GFP mRNA did not increase on stimulation of Jurkat cells. These results show that the decay rate of GFP mRNA due to the effect of haplotypes A1-A5 can be modulated by stimulation of Jurkat T cells.
The expression of GFP and its mRNA by transfected cells stimulated with aCD3/aCD28 was higher than in untreated cells (Fig. 4). However, a similar increase in GFP expression was common to all constructs, including control CMV-GFP (Fig. 4A). This was probably caused by a transcriptional effect of stimulation on the CMV promoter. In fact, the augmented expression of CMV-GFP caused by treatment with aCD3/aCD28 was at the mRNA level, without any change in its mRNA decay rate, indicating an increase in CMV promoter activity (Fig. 4, C and D; see also Fig. 3, A and C).
Protein binding properties of the TNFR2 3'-UTR
The results described above suggest the presence of cis-element(s) in the region of the TNFR2 3'-UTR fragment studied here that are able to increase reporter mRNA decay. To assess whether the TNFR2 3'-UTR is specifically recognized by cytosolic proteins, radiolabeled RNA probes corresponding to haplotypes A2 and A5 were generated, incubated with Jurkat cytosolic extracts, and analyzed by gel retardation assay. RNAs containing the A2 and A5 sequences were bound by Jurkat cytosolic proteins, generating protein-RNA complexes of similar mobility (Fig. 3E).
Haplotype A2 does not alter the transcriptional activity of the CMV promoter
The results described above indicate that the 3'-UTR of TNFR2 contains cis-element(s) able to augment reporter mRNA decay. To further evaluate whether there is a transcriptional contribution to the observed effect of haplotypes A1-A4, we performed nuclear run-on assays in Jurkat cells. Labeled transcripts were prepared from cells transfected with GFP-A2 or CMV-GFP and were hybridized to the probes shown in Fig. 4E. The expression of the GFP gene was similar in cells transfected with GFP-A2 or CMV-GFP (Fig. 4E), indicating similar transcription rates. Because the endogenous TNFR2 gene is silent in unstimulated Jurkat T cells (10, 52) (data not shown), a probe containing the TNFR2 3'-UTR fragment was used to determine the specificity of the assay. Of the two transcripts, GFP-A2 and CMV-GFP, only the former contains TNFR2 sequences. Accordingly, the signal detected with the TNFR2 probe in nuclear RNA prepared from cells transfected with CMV-GFP determines the background level of the assay (Fig. 4E).
These results suggest that the decrease in steady-state mRNA levels observed for GFP-A1, GFP-A2, GFP-A3, and GFP-A4 is due to posttranscriptional regulation, resulting in an enhanced rate of decay.
Discussion
Adenylate/uridylate-rich elements (AREs) are often found in the 3'-UTRs of posttranscriptionally regulatedmRNAs. Each ARE may represent a combination of structurally distinct domains, such as AU-rich motifs, U-rich elements, and poly(C) repeats, and the combination of these sequence elements determines its ultimate destabilizing function (53, 54, 55, 56, 57, 58). The TNFR2 3'-UTR contains no AU-rich motif. Instead, like the c-jun ARE, the representative of class III AREs, it contains numerous U/purine dinucleotides. GUUUG repeats like those present in TNFR2 3'-UTR also affect the stability of mRNAs (59, 60, 61, 62, 63, 64, 65). The GUUUG repeats in TNFR2 are close to the polymorphic region and are good candidates for functional cis-elements that control decay. Therefore, in this study a TNFR2 3'-UTR segment containing both the polymorphic region and GUUUG repeats was fused downstream of the GFP gene, and its influence on the stability of the chimeric transcript was assessed.
We have shown that the TNFR2 3'-UTR decreases GFP expression. Haplotypes A1, A2, A3, and A4 lowered GFP to a similar extent, more strongly than A5 (Fig. 2A). The mechanism used by A1-A4 to decrease reporter expression, compared with A5, could be a reduction in translation and/or stability of the GFP mRNA. The analysis of mRNA steady-state (Fig. 2G) points to a regulation of mRNA stability as its predominant target because significant differences were observed between transfectants A1-A4 and control CMV-GFP but not between A5 and CMV-GFP. The hypothesis that the mRNA decay rate is enhanced in A1-A4 transfectants is further sustained by the analysis of mRNA stability. Because the differences studied are small (1.5 times, Fig. 2G), it was not expected to observe extensive variation in the mRNA stability curve. In fact, a trend toward increased mRNA decay rates was found in A1-A4 (half-life 4.0 h), compared with A5 (half-life 5.8 h) (Fig. 3B). These data strengthened our previous conclusion that the A1-A4 haplotypes do decrease GFP expression more than A5, and their effect is most evident at the level of regulation of GFP mRNA steady-state.
TNFR2 transcript is stable (half-life > 27 h) in resting T cells but is destabilized in a stimulus-dependent manner (48). Consistent with the results of Raghavan et al. (48), we have shown that the instability conferred by 3'-UTR sequences increases on T-cell activation (Fig. 4). In both cases, cells transfected with A1-A4 or A5, showed increased mRNA decay rates, compared with control CMV-GFP (Fig. 3D). However, a trend toward higher stability of mRNA conferred by A5 is observed.
The decrease in steady-state GFP mRNA levels caused by the A1-A4 haplotypes and the fact that A5 does show a similar effect in stimulated cells suggests that a common determinant of instability is present in the TNFR2 3'-UTR. This could correspond to the U-rich cluster. In fact, the binding studies performed with the 3'-UTR fragment and Jurkat cell extracts demonstrate the existence of cellular factors that recognize the 3'-UTR of TNFR2, without appreciable differences between A5 and non-A5 alleles (Fig. 3E). Scanning this region to search for additional putative functional elements (66) revealed a CU-rich pattern related to 15-lipoxygenase-differentiation control element (67). The repeated CU-rich sequences were apposed in the secondary structure predicted for the 3'-UTR fragment (Fig. 5). These sequences could be recognized by K homology domain-containing RNA-binding proteins, which are able to control mRNA stability (68). The polymorphism 1690 T/C is close to the CU-rich sequences and could modulate the binding of factors to the CU-elements. However, both A2 and A5 contain a cytosine in position 1690 but conduce reporter expression to a different extent, suggesting that the contribution of nucleotide 1690 alone is not sufficient to determine the observed functional differences. Therefore, it is assumed that at least two polymorphisms act together in modulating trans-acting factor binding or accessibility to the U-rich cluster. Experiments aimed at supershifting the observed RNA-protein complexes with specific antibodies should help to characterize the proteins interacting with the TNFR2 3'UTR.
FIG. 5. Secondary structure of the TNFR2 3'-UTR 485-nt fragment. The predicted secondary structure of the TNFR2 RNA haplotype A5 is shown. The position of the three polymorphisms are circled dark. The location of CU-repeats is indicated by lines.
The phylogenetic conservation of the TNFR2 mRNA 3'-UTR is poor. Both mouse and human TNFR2 mRNAs contain long 3'-UTRs of more than 1000 nt. However, the region corresponding to the polymorphisms in the proximal 3'-UTR region is not well conserved, suggesting that this region is unlikely to contain prevalent cis-acting elements.
Polymorphisms in the 3'-UTR of the TNFR2 gene may contribute to the increased risk of Crohn’s disease (28), multiple sclerosis (69), and insulin resistance and obesity (27). Our results suggest that carriers of the common haplotypes A1-A4 will not show major differences in expression of cell membrane-associated TNFR2. However, A5 carriers may express increased levels of TNFR2.
Taken together, our results suggest that dysfunction of mRNA destabilizing signals that affect steady-state levels of TNFR2 mRNA could modify the abundance of TNFR2 mRNA in adipose tissue. We propose that an increased rate of TNFR2 mRNA decay protects cells from unrestrained TNF effects and that this protection will be diminished in A5 subjects (18). In a previous study, we found that younger subjects with the A5-allele (5.2% of nondiabetic subjects) show significantly higher body mass index (BMI) and leptin levels than subjects who do not carry this allele (see Ref. 27 , reported within A2 subjects). Of interest was the relationship between sTNFR2 levels and BMI and between sTNFR2 and leptin in those subjects in which the TNFR2 gene variants determined differences in BMI and leptin. Together with the present findings, we can speculate that increased BMI and leptin in A5 subjects is due to increased TNF effects. Consistent with this, levels of RNA transcripts encoding TNFR2 are significantly increased in the adipose tissues of obese diabetic animals. In muscle from these diabetic animals, RNA transcripts encoding both TNFR1 and TNFR2 were also significantly elevated, although the increase in TNFR2 transcript abundance was less marked than in fat. Overexpression of mRNA for TNF and both of its receptors could be at least partly normalized by treatment of the diabetic animals with the insulin-sensitizing agent pioglitazone (20).
Molecular characterization of proteins recognizing TNFR2 3'-UTR will help to determine the mechanism through which haplotypes differently regulate reporter mRNA decay. Furthermore, transfection analysis of mutants with deletions of the U-rich and C-rich elements will contribute to elucidating the activity of this putative ARE. We are currently performing additional studies to examine the functionality of these polymorphisms.
Acknowledgments
The authors are grateful to the flow cytometry team of the Serveis CientificoTècnics (Universitat de Barcelona) for help in these studies.
References
Vasalli P 1992 The pathophysiology of tumor necrosis factors. Annu Rev Immunol 10:411–452
Vandenabeele P, Declercq W, Beyaert R, Fiers W 1995 Two tumour necrosis factor receptors: structure and function. Trends Cell Biol 5:392–399
Ware CF, Crowe PD, Vanarsdale TL, Andrews JL, Grayson MH, Jerzy R, Smith CA, Goodwin RG 1991 Tumor necrosis factor (TNF) receptor expression in T lymphocytes. Differential regulation of the type I TNF receptor during activation of resting and effector T cells. J Immunol 147:4229–4238
Gehr G, Gentz R, Brockhaus M, Loetscher H, Lesslauer W 1992 Both tumor necrosis factor receptor types mediate proliferative signals in human mononuclear cell activation. J Immunol 149:911–917
Tartaglia LA, Goeddel DV, Reynolds C, Figari IS, Weber RF, Fendly BM, Palladino Jr MA 1993 Stimulation of human T-cell proliferation by specific activation of the 75-kDa tumor necrosis factor receptor. J Immunol 151:4637–4641
Zheng L, Fisher G, Miller RE, Peschon J, Lynch DH, Lenardo MJ 1995 Induction of apoptosis in mature T cells by tumour necrosis factor. Nature 377:348–351
Grell M, Becke FM, Wajant H, Mannel DN, Scheurich P 1998 TNF receptor type 2 mediates thymocyte proliferation independently of TNF receptor type 1. Eur J Immunol 28:257–263
Alexander-Miller MA, Derby MA, Sarin A, Henkart PA, Berzofsky JA 1998 Supraoptimal peptide-major histocompatibility complex causes a decrease in bc1–2 levels and allows tumor necrosis factor receptor II-mediated apoptosis of cytotoxic T lymphocytes. J Exp Med 188:1391–1399
Herbein G, Mahlknecht U, Batliwalla F, Gregersen P, Pappas T, Butler J, O’Brien WA, Verdin E 1998 Apoptosis of CD8+ T cells is mediated by macrophages through interaction of HIV gp120 with chemokine receptor CXCR4. Nature 395:189–194
Pimentel-Muinos FX, Seed B 1999 Regulated commitment of TNF receptor signaling: a molecular switch for death or activation. Immunity 11:783–793
Seckinger P, Isaaz S, Dayer JM 1989 Purification and biologic characterization of a specific tumor necrosis factor inhibitor. J Biol Chem 264:11966–11973
Engelmann H, Novick D, Wallach D 1990 Two tumor necrosis factor-binding proteins purified from human urine. Evidence for immunological cross-reactivity with cell surface tumor necrosis factor receptors. J Biol Chem 265:1531–1536
Schroder J, Stuber F, Gallati H, Schade FU, Kremer B 1995 Pattern of soluble TNF receptors I and II in sepsis. Infection 23:143–148
Gabay C, Cakir N, Moral F, Roux-Lombard P, Meyer O, Dayer JM, Vischer T, Guerne PA 1997 Circulating levels of tumor necrosis factor soluble receptors in systemic lupus erythematosus are significantly higher than in other rheumatic diseases and correlate with disease activity. J Rheumatol 24:303–308
Cope AP, Aderka D, Doherty M, Engelmann H, Gibbons D, Jones AC, Brennan FM, Maini RN, Wallach D, Feldmann M 1992 Increased levels of soluble tumor necrosis factor receptors in the sera and synovial fluid of patients with rheumatic diseases. Arthritis Rheum 35:1160–1169
Douni E, Kollias G 1998 A critical role of the p75 tumor necrosis factor receptor (p75TNF-R) in organ inflammation independent of TNF, lymphotoxin , or the p55TNF-R. J Exp Med 188:1343–1352
Hotamisligil GS 1999 Mechanisms of TNF--induced insulin resistance. Exp Clin Endocrinol Diabetes 107:119–125
Fernández-Real JM, Ricart W 2003 Insulin resistance and chronic cardiovascular inflammatory syndrome. Endocr Rev 24:278–301
Uysal KT, Wiesbrock SM, Hotamisligil GS 1998 Functional analysis of tumor necrosis factor (TNF) receptors in TNF--mediated insulin resistance in genetic obesity. Endocrinology 139:4832–4838
Hofmann C, Lorenz K, Braithwaite SS, Colca JR, Palazuk BJ, Hotamisligil GS, Spiegelman BM 1994 Altered gene expression for tumor necrosis factor- and its receptors during drug and dietary modulation of insulin resistance. Endocrinology 134:264–270
Kern PA, Saghizadeh M, Ong JM, Bosch RJ, Deem R, Simsolo RB 1995 The expression of tumor necrosis factor in human adipose tissue. Regulation by obesity, weight loss, and relationship to lipoprotein lipase. J Clin Invest 95:2111–2119
Hotamisligil GS, Arner P, Caro JF, Atkinson RL, Spiegelman BM 1995 Increased adipose tissue expression of tumor necrosis factor- in human obesity and insulin resistance. J Clin Invest 95:2409–2415
Saghizadeh M, Ong JM, Garvey WT, Henry RR, Kern PA 1996 The expression of TNF by human muscle. Relationship to insulin resistance. J Clin Invest 97:1111–1116
Hotamisligil GS, Arner P, Atkinson RL, Spiegelman BM 1997 Differential regulation of the p80 tumor necrosis factor receptor in human obesity and insulin resistance. Diabetes 46:451–455
Fernández-Real JM, Broch M, Ricart W, Casamitjana R, Gutierrez C, Vendrell J, Richart C 1998 Plasma levels of the soluble fraction of tumor necrosis factor receptor 2 and insulin resistance. Diabetes 47:1757–1762
Pantelidis P, Lympany PA, Foley PJ, Fanning GC, Welsh KI, du Bois RM 1999 Polymorphic analysis of the high-affinity tumor necrosis factor receptor 2. Tissue Antigens 54:585–591
Fernandez-Real JM, Vendrell J, Ricart W, Broch M, Gutierrez C, Casamitjana R, Oriola J, Richart C 2000 Polymorphism of the tumor necrosis factor- receptor 2 gene is associated with obesity, leptin levels, and insulin resistance in young subjects and diet-treated type 2 diabetic patients. Diabetes Care 23:831–837
Sashio H, Tamura K, Ito R, Yamamoto Y, Bamba H, Kosaka T, Fukui S, Sawada K, Fukuda Y, Tamura K, Satomi M, Shimoyama T, Furuyama J-I 2002 Polymorphisms of the TNF gene and the TNF receptor superfamily member 1B gene are associated with susceptibility to ulcerative colitis and Crohn’s disease, respectively. Immunogenetics 53:1020–1027
Glenn CL, Wang WY, Benjafield AV, Morris BJ 2000 Linkage and association of tumor necrosis factor receptor 2 locus with hypertension, hypercholesterolemia and plasma shed receptor. Hum Mol Genet 12:1943–1949
Geurts JM, Janssen RG, van Greevenbroek MM, van der Kallen CJ, Cantor RM, Bu X, Aouizerat BE, Allayee H, Rotter JI, de Bruin TW 2000 Identification of TNFRSF1B as a novel modifier gene in familial combined hyperlipidemia. Hum Mol Genet 9:2067–2074
Benjafield AV, Wang XL, Morris BJ 2001 Tumor necrosis factor receptor 2 gene (TNFRSF1B) in genetic basis of coronary artery disease. J Mol Med 79:109–115
Spotila LD, Rodriguez H, Koch M, Adams K, Caminis J, Tenenhouse HS, Tenenhouse A 2000 Association of a polymorphism in the TNFR2 gene with low bone mineral density. J Bone Miner Res 15:1376–1383
Peral B, San Millan JL, Castello R, Moghetti P, Escobar-Morreale HF 2002 The methionine 196 arginine polymorphism in exon 6 of the TNF receptor 2 gene (TNFRSF1B) is associated with the polycystic ovary syndrome and hyperandrogenism. J Clin Endocrinol Metab 87:3977–3983
van der Velden AW, Voorma HO, Thomas AA 2001 Vector design for optimal protein expression. Biotechniques 31:572, 574, 576–580
Waskiewicz AJ, Johnson JC, Penn B, Mahalingam M, Kimball SR, Cooper JA 1999 Phosphorylation of the cap-binding protein eukaryotic translation initiation factor 4E by protein kinase Mnk1 in vivo. Mol Cell Biol 19:1871–1880
Kar-Roy A, Dong W, Michael N, Li Y 2000 Green fluorescence protein as a transcriptional reporter for the long terminal repeats of the human immunodeficiency virus type 1. J Virol Methods 84:127–138
Espel E, García-Sanz JA, Aubert V, Menoud V, Sperisen P, Fernández N, Spertini F 1996 Transcriptional and translational control of TNF- gene expression in human monocytes by major histocompatibility complex class II ligands. Eur J Immunol 26:2417–2424
Kaabache T, Barraud B, Feldmann G, Bernuau D, Lardeux B 1995 Direct solution hybridization of guanidine thiocyanate-solubilized cells for quantitation of mRNAs in hepatocytes. Anal Biochem 232:225–230
Konarska MM, Sharp PA 1986 Electrophoretic separation of complexes involved in the splicing of precursors to mRNAs. Cell 46:845–855
Leibold EA, Munro HN 1988 Cytoplasmic protein binds in vitro to a highly conserved sequence in the 5' untranslated region of ferritin heavy- and light-subunit mRNAs. Proc Natl Acad Sci USA 85:2171–2175
Sambrook J, Russell DW 2001 Molecular cloning. In: Argentine J, ed. Molecular cloning: a laboratory manual. 3rd ed. New York: Cold Spring Harbor Laboratory Press; 17.23–17.29
Mathews DH, Sabina J, Zuker M, Turner DH 1999 Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J Mol Biol 288:911–940
Zuker M 2003 Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res 31:3406–3415
Forsberg EC, Zaboikina TN, Versaw WK, Ahn NG, Bresnick EH 1999 Enhancement of ?-globin locus control region-mediated transactivation by mitogen-activated protein kinases through stochastic and graded mechanisms. Mol Cell Biol 19:5565–5575
Magness ST, Tugores A, Brenner DA 2000 Analysis of ferrochelatase expression during hematopoietic development of embryonic stem cells. Blood 95:3568–3577
Raineri I, Wegmueller D, Gross B, Certa U, Moroni C 2004 Roles of AUF1 isoforms, HuR and BRF1 in ARE-dependent mRNA turnover studied by RNA interference. Nucleic Acids Res 32:1279–1288
Harrold S, Genovese C, Kobrin B, Morrison SL, Milcarek C 1991 A comparison of apparent mRNA half-life using kinetic labeling techniques vs decay following administration of transcriptional inhibitors. Anal Biochem 198:19–29
Raghavan A, Ogilvie RL, Reilly C, Abelson ML, Raghavan S, Vasdewani J, Krathwohl M, Bohjanen PR 2002 Genome-wide analysis of mRNA decay in resting and activated primary human T lymphocytes. Nucleic Acids Res 30:5529–5538
Mittrücker HW, Shahinian A, Bouchard D, Kündig TM, Mak TW 1996 Induction of unresponsiveness and impaired T cell expansion by staphylococcal enterotoxin B in CD28-deficient mice. J Exp Med 183:2481–2488
Saha B, Harlan DM, Lee KP, June CH, Abe R 1996 Protection against lethal toxic shock by targeted disruption of the CD28 gene. J Exp Med 183:2675–2680
Bachmann MF, McKall-Faienza K, Schmits R, Bouchard D, Beach J, Speiser DE, Mak TW, Ohashi PS 1997 Distinct roles for LFA-1 and CD28 during activation of naive T cells: adhesion versus costimulation. Immunity 7:549–557
Chan FK, Lenardo MJ 2000 A crucial role for p80 TNF-R2 in amplifying p60 TNF-R1 apoptosis signals in T lymphocytes. Eur J Immunol 30:652–660
Lagnado CA, Brown CY, Goodall GJ 1994 AUUUA is not sufficient to promote poly(A) shortening and degradation of an mRNA: the functional sequence within AU-rich elements may be UUAUUUA(U/A)(U/A). Mol Cell Biol 14:7984–7995
Xu N, Chen C-YA, Shyu A-B 1997 Modulation of the fate of cytoplasmic mRNA by AU-rich elements: key sequence features controlling mRNA deadenylation and decay. Mol Cell Biol 17:4611–4621
Czyzyk-Krzeska MF, Bendixen AC 1999 Identification of the poly(C) binding protein in the complex associated with the 3' untranslated region of erythropoietin messenger RNA. Blood 93:2111–2120
Paulding WR, Czyzyk-Krzeska MF 1999 Regulation of tyrosine hydroxylase mRNA stability by protein-binding, pyrimidine-rich sequence in the 3'-untranslated region. J Biol Chem 274:2532–2538
Tillmann-Bogush M, Heaton JH, Gelehrter TD 1999 Cyclic nucleotide regulation of PAI-1 mRNA stability. Identification of cytosolic proteins that interact with an A-rich sequence. J Biol Chem 274:1172–1179
Shao R, Shi Z, Gotwals PJ, Koteliansky VE, George J, Rockey DC 2003 Cell and molecular regulation of endothelin-1 production during hepatic wound healing. Mol Biol Cell 14:2327–2341
Shimizu N, Ohta M, Fujiwara C, Sagara J, Mochizuki N, Oda T, Utiyama H 1991 Expression of a novel immediate early gene during 12-O-tetradecanoylphorbol-13-acetate-induced macrophagic differentiation of HL-60 cells. J Biol Chem 266:12157–12161
You Y, Chen CY, Shyu AB 1992 U-rich sequence-binding proteins (URBPs) interacting with a 20-nucleotide U-rich sequence in the 3' untranslated region of c-fos mRNA may be involved in the first step of c-fos mRNA degradation. Mol Cell Biol 12:2931–2940
Levine TD, Gao F, King PH, Andrews LG, Keene JD 1993 Hel-N1: an autoimmune RNA-binding protein with specificity for 3' uridylate-rich untranslated regions of growth factor mRNAs. Mol Cell Biol 13:3494–3504
Chen CYA, Shyu AB 1994 Selective degradation of early-response-gene mRNAs: functional analyses of sequence features of the AU-rich elements. Mol Cell Biol 14:8471–8482
Peng H, Lever JE 1995 Regulation of Na(+)-coupled glucose transport in LLC-PK1 cells. Message stabilization induced by cyclic AMP elevation is accompanied by binding of a M(r) = 48,000 protein to a uridine-rich domain in the 3'-untranslated region. J Biol Chem 270:23996–24003
Peng SS, Chen CY, Shyu AB 1996 Functional characterization of a non-AUUUA AU-rich element from the c-jun proto-oncogene mRNA: evidence for a novel class of AU-rich elements. Mol Cell Biol 16:1490–1499
Paillard L, Legagneux V, Maniey D, Osborne HB 2002 c-Jun ARE targets mRNA deadenylation by an EDEN-BP (embryo deadenylation element-binding protein)-dependent pathway. J Biol Chem 277:3232–3235
Pesole G, Liuni S 1999 Internet resources for the functional analysis of 5' and 3' untranslated regions of eukaryotic mRNAs. Trends Genet 15:378
Ostareck-Lederer A, Ostareck DH, Standart N, Thiele BJ 1994 Translation of 15-lipoxygenase mRNA is inhibited by a protein that binds to a repeated sequence in the 3' untranslated region. EMBO J 13:1476–1481
Kiledjian M, Wang X, Liebhaber SA 1995 Identification of two KH domain proteins in the -globin mRNP stability complex. EMBO J 14:4357–4364
Ehling R, Gassner C, Lutterotti A, Strasser-Fuchs S, Kollegger H, Kristoferitsch W, Reindl M, Berger T 2004 Genetic variants in the tumor necrosis factor receptor II gene in patients with multiple sclerosis. Tissue Antigens 63:28–33(Irene Puga, Bego?a Lainez)