Skip Navigation

Cardiovascular Research 2003 59(1):160-168; doi:10.1016/S0008-6363(03)00296-7
© 2003 by European Society of Cardiology
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Lai, P. F.H
Right arrow Articles by Stewart, D. J
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lai, P. F.H
Right arrow Articles by Stewart, D. J
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Copyright © 2003, European Society of Cardiology

Downregulation of eNOS mRNA expression by TNF{alpha}: identification and functional characterization of RNA–protein interactions in the 3'UTR

Patrick F.H Laia,b,1, Farida Mohameda,1,2, Juan-Carlos Mongea and Duncan J Stewarta,b,*

aThe Terrence Donnelly Research Laboratories, Division of Cardiology, St. Michael's Hospital, Toronto, Ontario, Canada
bInstitute of Medical Science, University of Toronto, Toronto, Ontario, Canada

* Corresponding author. Room 7-081, Queen Wing, The Terrence Donnelly Heart Centre, Division of Cardiology, St. Michael's Hospital, 30 Bond Street, Toronto, Ontario, Canada M5B 1W8. stewartd{at}smh.toronto.on.ca

Received 28 January 2003; revised 19 February 2003; accepted 24 February 2003


   Abstract
Top
 Abstract
1. Introduction
 2. Methods
 3. Results
 4. Discussion
 References
 
Objective: We have previously shown that downregulation of endothelial nitric oxide synthase (eNOS) expression by tumour necrosis factor-{alpha} (TNF{alpha}) resulted entirely from the marked destabilization of the eNOS mRNA. As the 3'-untranslated region (3'UTR) in many eukaryotic mRNA has been well documented to bind regulatory trans-factors in the control of transcript stability, we have examined protein binding to this region of the eNOS mRNA. A high degree of homology amongst human and bovine 3'UTR also suggests that important functional features that are conserved through evolution are present within this region. Methods: RNA–protein interactions were studied in cross-linking assays, in which radiolabelled RNA encoding the human eNOS 3'UTR or selected sequences was incubated with cytoplasmic extracts of cultured human umbilical vein endothelial cells (HUVECs). Serial 5'- and 3'-truncated deletional mutations of the eNOS 3'UTR were generated to identify the specific binding sequences. eNOS mRNA expression in HUVECs was assessed by RT-PCR analysis. Results: Using radiolabelled RNA encoding the entire 418-nucleotide 3'UTR, we have identified ribonucleoprotein complexes (RNPs) of approximate molecular weights of 53, 56 and 66 kDa in the endothelial extracts. The formation of the 53- and 56-kDa RNPs was upregulated by TNF{alpha}, while the formation of the 66-kDa RNP was downregulated. Formation of the 53-kDa RNP was favoured by RNA fragments that contained sequences from the proximal and distal portions of the 3'UTR, whereas the formation of the 66-kDa RNP was favoured by RNA fragments with the AU-rich distal end. RNA fragments containing a CU-rich 158-nucleotide sequence from the medial portion of the eNOS 3'UTR (designated M158) favoured the formation of the 56-kDa RNP. Adenoviral gene transfer and overexpression of M158 RNA, as a protein-binding decoy to prevent the formation of the 56-kDa RNP on the endogenous transcripts, attenuated the TNF{alpha}-induced downregulation of eNOS mRNA in cultured endothelial cells. Conclusion: Our results demonstrate that the regulation of eNOS expression involves the specific binding of cytoplasmic proteins to highly conserved elements along the 3'UTR, and the 56-kDa RNP represents a novel regulatory trans-factor in the destabilization of eNOS transcripts.

KEYWORDS Gene expression; Endothelial factors; Nitric oxide synthase; Tumour necrosis factor {alpha}; RNA-binding factors


   1. Introduction
Top
 Abstract
 1. Introduction
2. Methods
 3. Results
 4. Discussion
 References
 
Endothelium-derived nitric oxide (NO) is an important mediator of vasodilation in response to a wide variety of stimuli, and represents a fundamental physiological mechanism maintaining functional and structural integrity of blood vessels. Release of NO in response to increases in blood flow and intimal shear stress is responsible for the moment-to-moment adjustment of vascular diameter, and maintenance of optimal conductance characteristics of the arterial tree [1–3]. Not only is endothelial NO a potent vasodilator, but it has important anti-platelet and anti-proliferative actions which, over the long-term, may contribute to vascular homeostasis by preventing thrombus formation and abnormal growth of vascular cells within the intima [3]. Thus, the dysregulation of NO synthesis in the endothelium may have far ranging influences in vascular pathobiology.

Endothelial cells express a ‘constitutive’ NO-synthase (eNOS) [4,5] which can be profoundly downregulated in the presence of cytokines, including tumour necrosis factor-{alpha} (TNF{alpha}) [6,7], as well as by pathophysiological stimuli such as hypoxia [8] and endothelial cell proliferation [9]. Downregulation of eNOS is likely an important mechanism resulting in endothelial dysfunction, and contributing to a variety of human vascular disorders characterized by reduced endothelium-dependent dilation and impaired ability of the endothelium to generate NO [10].

The downregulation of eNOS induced by TNF{alpha} was demonstrated by our group [6] as well as Yoshizumi et al. [7] to be due to the selective destabilization of its mRNA, with no alteration in RNA transcriptional activity [7]. Similarly, reduced steady-state levels of eNOS expression in hypoxia [8] and in rapidly proliferating endothelial cells [9] resulted from increased mRNA degradation. In contrast, transcript stabilization is involved in the upregulation of eNOS by statins [11,12], vascular endothelial growth factor [13] and estrogen [14].

The aim of the present study was to determine and characterize the molecular mechanisms controlling eNOS expression and mRNA stability. As the 3'-untranslated region (3'UTR) is considered to have paramount importance in the regulation of stability of a variety of eukaryotic messages [15–18], we specifically examined RNA–protein interactions at this region of the eNOS mRNA.


   2. Methods
Top
 Abstract
 1. Introduction
 2. Methods
3. Results
 4. Discussion
 References
 
2.1 Cell culture and preparation of cellular extracts
Human umbilical vein endothelial cells (HUVECs) obtained from the American Type Culture Collection were cultured in Ham's F12 medium, supplemented with 15% fetal calf serum, 100 U/ml penicillin, 100 µg/ml streptomycin, 100 µg/ml heparin (all from Invitrogen), and 20 µg/ml endothelial cell growth factor (Roche). Prior to the preparation of total RNA and/or cytoplasmic extracts, confluent HUVECs were incubated overnight with or without TNF{alpha} (Sigma) at 100 U/ml added to the culture medium, unless otherwise stated. Total RNA was extracted using either the Trizol Reagent (Invitrogen) or the GenElute RNA Kit (Sigma), following the manufacturers’ protocols. Cytoplasmic extracts were prepared from cells in a hypotonic lysis buffer consisting of 10 mM Hepes (pH 7.9), 40 mM KCl, 3 mM MgCl2, 1 mM DTT, 5% glycerol, 0.2% Nonidet P-40, 1 µg/ml aprotinin, 0.5 µg/ml leupeptin, and 0.5 mM phenylmethylsulfonylfluoride (all from Sigma) and mechanically homogenized. The nuclei were removed by centrifugation (4500xg, 15 min, 4°C), and the cytoplasmic fraction (supernatant) was aliquoted and immediately frozen at –80°C. Protein content was determined by Bradford's method (BioRad).

2.2 Cloning of heNOS 3'UTR and mutants
A cDNA fragment corresponding to the entire 3'UTR of heNOS (418 bases, Fig. 1) was amplified by reverse transcription-polymerase chain reaction (RT-PCR) from total RNA extracted from HUVECs. The PCR product was ligated into the pGEM-T Easy vector (Promega), sequenced and found to contain the identical 418-bp 3'UTR sequence as published for human eNOS in Genbank (Accession number: L10709 [GenBank] ). Using this cloned cDNA fragment as the template, a variety of deletional mutants of the heNOS 3'UTR (Fig. 2F) were generated by PCR and ligated into the pCR2.1-TOPO vector (Invitrogen). The sense PCR primers included an SP6 RNA polymerase site immediately upstream of the desired eNOS 3'UTR sequence for in vitro transcription, whereas the antisense primers contained an SspI site immediately downstream for plasmid linearization. These truncational mutants were named according to the length (in nucleotides) of the 3'UTR sequence included, following a ‘P’ or a ‘D’ to designate that they contained either the ‘P’roximal or ‘D’istal end of the eNOS 3'UTR, respectively. The mutant M158 contained the medial 158 bases of the eNOS 3'UTR, a region excluded by the proximal mutant P147 and the distal mutant D113. The mutants MD86 and MD39, which contained the distal 86 and 39 nucleotides in the M158 region, were generated by the same PCR-based strategy. A series of engineered concatamers, namely (CCUCU)10, (UUCUC)10, (CU)25, (CUUU)12 and (AUUU)12 were also generated by PCR using a long oligonucleotides with the desired concatamer sequence, flanked by two short sequences of restriction sites, as a template, and short oligonucleotides complementary to the flanking restriction sequences as primers. Amplified DNA fragments were ligated into pGEM-3Z at the EcoRI and HindIII sites. All plasmid constructs were sequenced to confirm their base sequence identity.


Figure 1
View larger version (31K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 A comparison of the human eNOS 3'UTR RNA sequence (top) and the bovine equivalent (bottom). Vertical lines indicate homology between the two sequences. Boxes indicate Domains X, Y and Z, i.e., regions with at least 90% identity for 20 or more consecutive bases. The sequences in boldface indicate previously reported and novel cis-acting motifs in the regulation of RNA stability: CCUCC, CCUCU (underlined) and AUUUA (underlined and italicized).

 

Figure 2
View larger version (43K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 RNA–protein cross-linking experiments using HUVEC cytoplasmic extracts with radiolabelled RNA encoding eNOS 3'UTR and truncated mutants. (A) A representative experiment using radiolabelled 3'UTR RNA in the absence (lane 1) or presence (lane 2) of HUVEC extracts. (B,C) Cross-linking studies using the various truncational mutants of the eNOS 3'UTR. (D) A representative competition study using radiolabelled D132 RNA against 100-molar excess of unlabelled P305 RNA (lane 3) or unlabelled D132 RNA (lane 4). (E) Cross-linking experiments using 15 fmol of 3'UTR RNA (lane 1) and 50 fmol of RNA concatamers (lanes 2–5). (F) Schematic representation of the eNOS 3'UTR and truncational mutants, and a summary of the binding pattern of the 53-, 56-, and 66-kDa factors. –, no detectable bands; +/–, barely detectable or weak bands; +, clear bands; ++, strong bands; +++, intense bands.

 
2.3 In vitro transcription
The vectors containing the eNOS 3'UTR constructs were linearized by SspI digestion prior to being used as templates for in vitro transcription, except that the 3'-truncated mutants P219 and P181 were generated by restriction digestion of P305 plasmid with DraII and StuI, respectively. Plasmid templates for the concatamers were linearized with EcoRI or HindIII, depending on the orientation of the insert. SP6 or T7 RNA polymerases were used in the in vitro transcription reactions according to the supplier's protocol (Promega), with or without 50 µCi of [32P]UTP or [32P]CTP (Amersham). A small sample of each [32P]RNA was electrophoresed and autoradiographed to confirm the size and integrity of the RNA fragments before RNA-binding experiments were performed.

2.4 Ultra-violet (UV) irradiation cross-linking studies
Fifteen µg of HUVEC cytoplasmic protein preparations were mixed with 300–500 ng of total yeast RNA (non-specific competitor) or a 100-fold molar excess of identical unlabelled (cold) RNA (specific competitor) in a reaction mixture containing 20 mM Hepes (pH 7.6), 60 mM KCl, 1 mM MgCl2 and 10% glycerol and chilled on ice. Then, 15 fmol of in vitro transcribed [32P]RNA was added to make a total volume of 20 µl for further incubation at 37°C for 10 min. RNA–protein interactions were then stabilized by ultra-violet irradiation in a GS GeneLinker UV Chambre (BioRad), after which unbound RNA fragments were broken down by RNases A and T1 (Roche) at 37°C for 15 min. Each reaction was then mixed with SDS sample buffer with or without mercaptoethanol, boiled for and resolved in an SDS–polyacrylamide electrophoretic gel. Gels were dried and then imaged either by autoradiography or using a Cyclone phosphorimager (Canberra-Packard) to allow densitometric quantitation of band intensities using the MolecularAnalyst software (v1.5, BioRad). Experiments with each individual RNA fragment were repeated at least five times with similar results.

2.5 Replication-deficient recombinant adenoviral infections and RT-PCR
Recombinant adenoviruses engineered to carry the M158 fragment (AdV-M158) or the control β-galactosidase cDNA (AdV-LacZ) for constitutive expression under the control of the cytomegalovirus (CMV) promoter were kindly provided by Drs. Karen F. Kozarsky, Mary H. Donahee and Eliot H. Ohlstein (Cardiovascular Biology, GlaxoSmithKline, King of Prussia, PA). Confluent HUVECs grown in 60-mm plates were infected at approximately 3x104 adenoviral particles/cell for 2 h in culture medium containing reduced serum (1%). No apparent cytopathic effects were observed. One to 3 days after each adenoviral infection experiment, one plate of cells infected to overexpress LacZ was fixed and stained with X-gal for β-galactosidase activity to confirm a near 100% gene transduction efficiency. After 2–4 days post-infection, cells were treated with 100 U/ml of TNF{alpha} for 9–12 h and then lysed for total RNA extraction.

For end-stage RT-PCR, each 20-µl reverse transcription (RT) was performed on 1 µg of RNA with random hexamers and Moloney murine leukemia virus reverse transcriptase (MMLV-RT, Invitrogen) following the manufacturer's instructions. To control for possible genomic or adenoviral DNA contaminations, RTs without MMLV-RT were performed in parallel. Two µl of each RT were then used in 20-µl PCRs with 1 unit of Taq polymerase (Pharmacia or Roche) to amplify DNA fragments representing the exogenous M158 (sense primer: 5'-GAG CAT TCG CCC TTC AGA TTT AGG TGA CA-3', antisense primer: 5'-GAC CGC GGA ATA TTG AGA GAG GCA AGA GGA AT-3', fragment size: 211 bp) and endogenous GAPDH (sense primer: 5'-CTC TAA GGC TGT GGG CAA GGT CAT-3', antisense primer: 5'-GAG ATC CAC CAC CCT GTT GCT GTA-3', fragment size: 343 bp). PCR products were electrophoresed in 2% agarose and stained with ethidium bromide for visualization.

For real-time quantitative RT-PCR, 0.1 µg of RNA and Omniscript reverse transcriptase (Qiagen) were used in RTs, which were subsequent diluted 1:5 in PCR-grade water. Four µl of each diluted RT was then used in 20-µl PCRs in duplicates to assess the RNA abundance of eNOS (sense primer: 5'-GTG GCT GTC TGC ATG GAC CT-3', antisense primer: 5'-CCA CGA TGG TGA CTT TGG CT-3') and β-actin (sense primer: 5'-AGC CTC GCC TTT GCC GA-3', antisense primer: 5'-CTG GTG CCT GGG GCG-3', Ref. [19]) using the SYBR Green PCR Master Mix and the ABI PRISM 7900HT sequence detection system (Applied Biosystems). RTs and PCRs with various dilutions of total HUVEC RNA were done in parallel to determine the amplification efficiencies (E) according to the equation:

Formula
[20,21], where the slope is that of the standard curve plot (threshold cycle or crossing point, CP, versus RNA input) generated by the ABI PRISM system. Relative eNOS RNA expression was then calculated in comparison to β-actin RNA expression using Pfaffl's formula:

Formula
[21]. The relative abundance of eNOS RNA in control cells not treated with TNF{alpha} was expressed as 100% for comparative purposes.


   3. Results
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
4. Discussion
 References
 
Sequence analysis of the human eNOS 3'UTR, performed to identify evolutionarily conserved sequences that may have functional significance, revealed a high level of homology at approximately 66% with the bovine eNOS 3'UTR, particularly along the distal half of the 3'UTR. As shown in Fig. 1, three domains of 20 or more nucleotides exhibiting near identity (≥90%) between the human and bovine sequences were identified. Domain X (nucleotides 211–251 in the heNOS 3'UTR) is C-rich and contains one conserved copy of the CCUCC motif, which was previously reported to be the binding sequence for a 39-kDa ribonucleoprotein (RNP) complex on the human {alpha}-globin mRNA [22]. Domain Y (283–304) is CU-rich and contains two copies of a CCUCU motif. Domain Z (339–401) is AU-rich and carries two conserved copies of the AUUUA motif commonly found in the 3'UTR of labile messages [15,16].

Cross-linking experiments were performed with in vitro transcribed and radiolabelled RNA fragments encoding the complete heNOS 3'UTR and various truncational mutants (Fig. 2). When radiolabelled 3'UTR was incubated with HUVEC cytoplasmic extracts, multiple bands were observed, corresponding to RNP complexes of approximate molecular weights of 53-, 56- and 66-kDa (Fig. 2A). Bands or complexes at 45 and 100 kDa were not consistently observed and therefore not subjected to further investigation. No distinct bands could be observed from control reactions performed without the cytoplasmic extracts, demonstrating that the bands were not attributable to incompletely digested RNA fragments (Fig. 2A). The binding data from experiments using the various heNOS 3'UTR RNA fragments are shown in Fig. 2B and C, and summarized in Fig. 2F. The proximal mutant P305, which lacked the distal Domain Z, displayed binding to the 53- and 56-kDa RNP, but not the 66-kDa RNP (Fig. 2B, lane 1). P267, a shorter proximal mutant lacking both Domains Y and Z, also exhibited binding to the 53- and 56-kDa RNP and not the 66-kDa RNP, but binding to the 56-kDa RNP was weaker (lane 2). Smaller proximal mutants that do not contain any of the three conserved domains displayed binding to the 53-kDa RNP only (lanes 3–5). The distal mutants D285, D271 and D132 (lanes 8–10) displayed the same pattern of binding as the intact 3'UTR, giving strong signals at 53, 56 and 66 kDa. The D113 mutant, which lacked Domains X and Y but retained the AU-rich Domain Z, formed RNP complexes at 66 and 53 kDa, but not at 56 kDa (lane 11). The medial mutant M158, carrying conserved Domains X and Y, showed the formation of the 56-kDa RNP complex alone (lane 6). Deletional constructs containing distal portions of the M158 region: MD86 and MD39 were used to further determine the binding site of this 56-kDa RNP (Fig. 2C). MD39, which contained the distal 39 nucleotides of M158 and included the two CCUCU motifs in Domain Y, was sufficient to form the 56-kDa RNP complex (lane 3).

Competition studies with unlabelled RNA fragments were performed to determine the specificity of protein binding. Fig. 2D shows a representative experiment using radiolabelled D132 RNA fragments. In the absence of a specific RNA competitor, clear bands representing the 53-, 56- and 66-kDa complexes were observed (lane 2). Upon competition with a 100-molar excess of unlabelled D132 RNA (lane 4), no discernible binding could be detected. A 100-molar excess of unlabelled P305 RNA (lane 3), a proximal fragment which lacks Domain Z but overlaps with D132 in Domain Y, competed for the 53- and 56-kDa RNP complexes, but not for the 66-kDa RNP.

Further cross-linking experiments with the [32P]UTP-labelled synthetic concatamers revealed that the 56-kDa RNP complex could be formed with various CU-rich RNA motifs. As shown in Fig. 2E, a strong band at 56 kDa was observed with the RNA concatamers (CCUCU)10 (10 tandem repeats of the CCUCU motif), (UUCUC)10 and (CUUU)12, whereas the concatamer (CU)25 exhibited very weak binding activity (data not shown). The same pattern of binding was obtained when [32P]CTP-labelled concatamers were used (data not shown). In contrast, the concatamer (AUUU)12 formed several RNP complexes of different sizes, appearing as multiple bands in the autoradiogram (lane 5).

A representative cross-linking experiment is shown in Fig. 3 to demonstrate the effect of TNF{alpha} on RNA–protein interactions. In cytoplasmic extracts from HUVECs treated overnight with 100 U/ml of TNF{alpha}, the binding of the 66-kDa RNP to radiolabelled 3'UTR RNA fragments was 42% less than that in extracts from untreated control cells (P<0.05, n = 4). In contrast, the binding of the 56- and 53-kDa RNP was increased in extracts from TNF{alpha}-treated HUVECs by 28 and 24%, respectively (P<0.01, n = 8).


Figure 3
View larger version (22K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3 Effect of TNF{alpha} on RNA–protein interactions. (A) A representative RNA–protein cross-linking experiment using radiolabelled eNOS 3'UTR RNA fragments with cytoplasmic proteins from HUVECs extracted under control (untreated) conditions and under TNF{alpha} treatment, 100 U/ml overnight. (B) Graphical summary data of RNA–protein binding, reflecting percentage change (from control) of densitometric band intensity representing the three RNA-binding factors under TNF{alpha}-treated conditions. Mean±S.E.M.; n = 4 for 66-kDa RNP, and n = 8 for 56- and 53-kDa RNP; *P<0.05, **P><0.01 by unpaired t-tests.

 
In order to determine the functional importance of the formation of the 56-kDa RNP on the M158 region in the heNOS 3'UTR RNA, which can be upregulated by TNF{alpha}, HUVECs were transduced with an adenoviral infection to overexpress the M158 sequence as an ‘exogenous RNA decoy’ designed to compete for protein binding against the endogenous eNOS transcripts. After 2–4 days post-infection, cells were then treated with TNF{alpha} (100 U/ml) for 9–12 h and total RNA was extracted for RT-PCR analyses. The expression of the adenoviral-derived M158 RNA in infected cells was confirmed by end-stage RT-PCR using primers that would amplify vector-derived sequences of the exogenous RNA but not the M158 region of the endogenous eNOS transcript (Fig. 4A). The expression of eNOS RNA was then determined by real-time RT-PCR. As shown in Fig. 4B, the eNOS RNA level was similarly downregulated by TNF{alpha} in mock-infected cells (white column) and cells over-expressing LacZ after a control adenovirus infection (gray column), eNOS RNA expression in the presence of TNF{alpha} relative to untreated controls were only 26.9±3.6 and 29.6±3.9%, respectively (n = 13, not significantly different). This TNF{alpha}-mediated downregulation of eNOS transcript levels was attenuated in cells expressing the adenoviral-derived M158 RNA (black column), resulting in a significantly higher relative eNOS RNA expression at 41.5±6.3% (n = 13, P<0.01 in paired t-tests against mock-infected or AdV-LacZ-infected controls) after the same TNF{alpha} treatment.


Figure 4
View larger version (37K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4 RT-PCR analysis of RNA abundance. (A) A representative 2% agarose electrophoretic gel, stained with ethidium bromide to visualize amplified DNA fragments of adenoviral-derived M158 and endogenous GAPDH sequences. RT-PCR was performed using RNA extracted from endothelial cells that had been mock-infected (lanes 1–4), infected with AdV-M158 to overexpress the M158 region, the putative binding sequence for the 56-kDa RNA–protein complex on the eNOS 3'UTR (lanes 6–13) or infected by AdV-LacZ to overexpress β-galactosidase as a control (lanes 15–18). Lanes 10–13 contain negative control reactions in which MMLV-RT was excluded in RT, thus no cDNA was synthesized and only contaminating adenoviral DNA or genomic DNA, if any, would be amplified. Lanes 5 and 14 contain DNA size markers. (B) Relative eNOS RNA expression in the presence of TNF{alpha} (100 U/ml), as determined by real-time PCR, normalized to β-actin RNA levels and expressed as a percentage of eNOS RNA levels in untreated (TNF{alpha}-free) controls. TNF{alpha}-mediated downregulation of eNOS RNA was attenuated in cells overexpressing M158 RNA as a ‘protective decoy’, resulting in a significantly higher level of eNOS RNA than in both mock-infected and AdV-LacZ-infected cells after the same TNF{alpha} treatment. Mean±S.E.M., n = 13, *P<0.01 by paired t-tests.

 

   4. Discussion
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
References
 
Post-transcriptional mechanisms play an important role in the regulation of expression of an increasing number of eukaryotic genes, with specific elements involved in the regulation of RNA stability found in the 3'UTR of many messages [15–18]. Analysis of sequence homology between human and bovine eNOS 3'UTR led to the identification of three candidate regulatory elements in highly conserved domains on the mRNA (Fig. 1). Domain X (nucleotides 211–251 of the heNOS 3'UTR) was C-rich, containing a CCUCC motif previously reported in the human {alpha}-globin mRNA to be a cis-acting element that contributes to the stabilization of the transcript and increases {alpha}-globin expression [22]. Domain Y (283–304) was a CU-rich region containing two copies of a CCUCU motif, a motif not previously described as a regulatory sequence. Domain Z (339–401), an AU-rich sequence, contained two AUUUA motifs previously identified in the 3'UTR of many rapidly degraded mRNA, such as c-fos, c-jun, and c-myc [15,16].

In the present study, three specific RNA-binding factors that bind to the human eNOS 3'UTR RNA have been identified in HUVEC cytoplasmic extracts. The binding of the three factors varied differentially in response to TNF{alpha}. However, at this point, it is not known if the changes in binding reflected changes in the binding affinity or the relative abundance of individual RNA-binding factors. The 66-kDa factor, the largest of the three RNPs, bound to the AU-rich distal part of the heNOS 3'UTR, as all truncation RNA mutants lacking this sequence failed to form this RNP complex (Fig. 2B, lanes 1–6). The failure of P305 RNA to compete with D132 RNA for binding to the 66-kDa factor (Fig. 2D) also supported that this factor binds to sequences distal to the P305 region. The binding site was likely located in the conserved region in Domain Z, which contained two AUUUA motifs. Cross-linking studies using (AUUU)12 RNA concatamers revealed the presence of multiple AUUUA-binding factors in endothelial cytoplasmic fractions (Fig. 2D, lane 5), including one at 66 kDa. RNA-binding proteins at this size and with an affinity for the AUUUA motif or AU-rich sequences have not been previously reported, thus, this 66-kDa RNP may represent a novel AU-binding factor involved in RNA regulation. It is likely a transcript stabilization factor, as the reduction in binding to eNOS 3'UTR RNA in response to TNF{alpha} (Fig. 3) correlated well with TNF{alpha}-mediated mRNA downregulation (Fig. 4).

The 53-kDa RNP could be formed with truncation mutants carrying either the proximal or the distal end of heNOS 3'UTR in the presence of HUVEC cytoplasmic extracts (Fig. 2B). Although the medial M158 region can be excluded, the precise binding site of this factor could not be determined. The 53-kDa factor may bind to Domain Z along with the 66-kDa factor, or it may bind promiscuously to other less conserved sequences at either ends of the 3'UTR. A 51-kDa bovine aortic endothelial protein was previously described to bind selectively to a 43-nucleotide GC-rich fragment of the proximal beNOS 3'UTR immediately downstream of the stop codon [23]. Our 53-kDa factor may be a related human protein with lesser binding specificity.

The 56-kDa RNP can be formed in the presence of HUVEC cytoplasmic extracts with the CU-rich M158 mutant (Fig. 2B, lane 6), and indiscriminately with RNA concatamers with tandem repeats of CU-rich motifs (Fig. 2E, lanes 2–4), suggesting that this 56-kDa binding factor has an affinity for RNA rich in C and U. The binding site was likely confined within the medial region of the heNOS 3'UTR, as neither mutants P147 nor D113 could form the 56-kDa RNP (Fig. 2B, lanes 5 and 11). Excess cold P305 RNA effectively competed with radiolabelled D132 RNA for the 56-kDa RNP but not for the 66-kDa RNP (Fig. 2D, lane 3), thus the CU-rich Domain Y, where the two mutant RNA fragments overlap, is very likely to be a site of RNA–protein interaction for the 56-kDa factor. MD39, a short 39-nucleotide RNA fragment containing Domain Y and the flanking sequences was sufficient to form the 56-kDa RNP in the presence of HUVEC cytoplasmic extracts (Fig. 2C, lane 3), strongly suggesting that Domain Y is indeed the binding sequence for the 56-kDa factor. This was also supported by the reduced band intensity of the 56-kDa RNP formed by mutant P267 compared to that formed by mutant P305, a longer proximal fragment that includes Domain Y. Further serial truncation to remove Domain X led to the further loss of binding to the 56-kDa RNP on the shorter proximal mutants P219, P181 and P147 (Fig. 2B, lanes 3–5). This suggests that the C-rich Domain X is possibly a second or secondary site for this RNA–protein interaction. However, the relative importance of Domains X and Y in the binding of the 56-kDa factor to the M158 RNA requires further investigation.

The functional importance of the RNA–protein interaction between the 56-kDa factor and the heNOS transcript was suggested by the increase in binding after treatment with TNF{alpha} (Fig. 3), which downregulates eNOS mRNA expression by transcript destabilization [6]. Moreover, overexpression of RNA carrying the binding sequences of the 56-kDa factor (the M158 region) as an RNA decoy designed to compete for the binding with the endogenous heNOS transcript led to the attenuation of TNF{alpha}-mediated downregulation of the heNOS mRNA levels (Fig. 4). These results suggest that the binding of the 56-kDa factor to heNOS mRNA was required in the TNF{alpha}-induced mRNA downregulation. The implications are that this 56-kDa CU-binding protein may be a functional trans-factor for transcript destabilization, and the M158 region the corresponding cis-element in the regulation of eNOS expression in cultured human endothelial cells. However, the identity of this 56-kDa factor remains unknown. A 56-kDa polypyrimidine tract binding protein (PTB) has recently been reported to bind to CU-rich sequences in the 3'UTR of rat insulin mRNA in pancreatic cells [25] and the murine inducible NOS mRNA in liver cells [26]. Thus, we speculate that our 56-kDa human endothelial NOS mRNA-binding factor with an affinity for CU-rich 3'UTR sequences could be related to PTB, although this protein targets the inducible rather than endothelial NOS isoform.

The 56-kDa human eNOS mRNA-binding factor is quite similar and possibly homologous to a 60-kDa bovine protein associated with CU-rich sequences in the bovine eNOS mRNA, as previously reported by Gonzalez-Fernandez et al. [24] The mRNA-binding of the 60-kDa bovine factor was found to be dose-dependently reduced by cerivastatin in a manner that correlates to the concomitant increases in eNOS mRNA abundance, suggesting but not establishing a causal relationship between the RNA–protein interaction and eNOS expression [24]. In the current study, the overexpression of M158 RNA attenuated the eNOS downregulation, confirming the importance of the 56-kDa trans-factor and the M158 RNA cis-element in the regulation of eNOS expression in human endothelial cells.

Although the overexpression of M158 RNA has only led to a modest increase in eNOS RNA abundance in cultured endothelial cells, our results have clearly demonstrated the utility of this molecular strategy in modulating eNOS expression. Given the importance of NO in endothelial biology, the M158 RNA overexpression could be used to enhance eNOS gene expression, countering endothelial dysfunction and other vascular pathologies related to reduced NO production. The modulation of eNOS expression in the vasculature by reducing or even preventing the loss of eNOS mRNA could be an attractive alternative to NOS gene transfer in future clinical applications of molecular interventions.

In summary, we have identified three cytoplasmic factors with strong binding affinity for highly conserved RNA sequences in the human eNOS 3'UTR as candidate regulatory trans-factors and cis-elements in the control of eNOS expression in cultured human endothelial cells. The binding of trans-acting cytoplasmic factors to putative regulatory cis-elements within the eNOS 3'UTR, and the modulation of these interactions by TNF{alpha}, strongly suggest a role in the post-transcriptional regulation of RNA stability. Overexpression of RNA sequences designed to disrupt the RNA–protein interactions between the 56-kDa factor and the endogenous eNOS message exhibited a protective effect against TNF{alpha}-mediated downregulation of eNOS expression, clearly demonstrating a critical role in the eNOS gene regulation. These results provide a basis for future studies on the mechanisms of regulation of human eNOS mRNA stability, and suggest potential molecular targets for the manipulation of eNOS expression in endothelial cells for therapeutic purposes.

Time for primary review 14 days.


   Acknowledgements
 
The authors would like to thank Drs. Eliot H. Ohlstein, Karen F. Kozarsky and Mary H. Donahee (Cardiovascular Biology, GlaxoSmithKline, King of Prussia, PA) for providing the adenoviruses used in this study, Drs. Gerald A. Proteau, Neil Fam and Phyllis Billia for providing primers and assistance with real-time PCR, as well as Mr. Kevin Wyllie and Dr. JianXin Ren for their technical expertise in RNA–protein binding experiments. P.F.H. Lai was a recipient of a doctoral studentship from the Medical Research Council of Canada. This investigation was funded by the Canadian Institutes for Health Research and the Heart and Stroke Foundation of Ontario.


   Notes
 
1 These authors contributed equally in this study. Back

2 Room 1149-1, Henry F. Hall Building, 1455 De Maisonneuve Blvd. West, Department of Chemistry & Biochemistry, Concordia University, Montréal, Québec, Canada H3G 1M8. Back


   References
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
 References
 

  1. Griffith T.M., Edwards D.H., Davies R.L., Harrison T.J., Evans K.T. EDRF coordinates the behaviour of vascular resistance vessels. Nature (1987) 329:442–445.[CrossRef][Medline]
  2. Furchgott R.F., Vanhoutte P.M. Endothelium-derived relaxing and contracting factors. FASEB J (1989) 3:2007–2018.[Abstract]
  3. Moncada S., Palmer R.M., Higgs E.A. Nitric oxide: physiology, pathophysiology, and pharmacology. Pharmacol Rev (1991) 43:109–142.[Web of Science][Medline]
  4. Marsden P.A., Schappert K.T., Chen H.S., Flowers M., Sundell C.L., Wilcox J.N., Lamas S., Michel T. Molecular cloning and characterization of human endothelial nitric oxide synthase. FEBS Lett (1992) 307:287–293.[CrossRef][Web of Science][Medline]
  5. Nishida K., Harrison D.G., Navas J.P., Fisher A.A., Dockery S.P., Uematsu M., Nerem R.M., Alexander R.W., Murphy T.J. Molecular cloning and characterization of the constitutive bovine aortic endothelial cell nitric oxide synthase. J Clin Invest (1992) 90:2092–2096.[Web of Science][Medline]
  6. Mohamed F., Monge J.C., Gordon A., Cernacek P., Blais D., Stewart D.J. Lack of role for nitric oxide (NO) in the selective destabilization of endothelial NO synthase mRNA by tumor necrosis factor-{alpha}. Arterioscler Thromb Vasc Biol (1995) 15:52–57.[Abstract/Free Full Text]
  7. Yoshizumi M., Perrella M.A., Burnett J.C. Jr., Lee M.E. Tumour necrosis factor downregulates an endothelial nitric oxide synthase mRNA by shortening its half-life. Circ Res (1993) 73:205–209.[Abstract]
  8. McQuillan L.P., Leung G.K., Marsden P.A., Kostyk S.K., Kourembanas S. Hypoxia inhibits expression of eNOS via transcriptional and posttranscriptional mechanisms. Am J Physiol (1994) 267:H1921–H1927.[Web of Science][Medline]
  9. Flowers M.A., Wang Y., Stewart R.J., Patel B., Marsden P.A. Reciprocal regulation of endothelin-1 and endothelial constitutive NOS in proliferating endothelial cells. Am J Physiol (1995) 269:H1988–H1997.[Web of Science][Medline]
  10. Vallance P., Collier J. Biology and clinical relevance of nitric oxide. Br Med J (1994) 309:453–457.[Free Full Text]
  11. Laufs U., La Fata V., Liao J.K. Inhibition of 3-hydroxy-3-methylglutaryl (HMG)-CoA reductase blocks hypoxia-mediated downregulation of endothelial nitric oxide synthase. J Biol Chem (1997) 272(50):31725–31729.[Abstract/Free Full Text]
  12. Laufs U., La Fata V., Plutzky J., Liao J.K. Upregulation of endothelial nitric oxide synthase by HMG CoA Reductase Inhibitors. Circulation (1998) 97:1129–1135.[Abstract/Free Full Text]
  13. Bouloumie A., Schini-Kerth V.B., Busse R. Vascular endothelial growth factor upregulates nitric oxide synthase expression in endothelial cells. Cardiovasc Res (1999) 41:773–780.[Abstract/Free Full Text]
  14. Sumi D., Hayashi T., Jayachandran M., Iguchi A. Estrogen prevents destabilization of endothelial nitric oxide synthase mRNA induced by tumour necrosis factor {alpha} through estrogen receptor mediated system. Life Sci (2001) 69:1651–1660.[CrossRef][Web of Science][Medline]
  15. Jackson R.J. Cytoplasmic regulation of mRNA function: the importance of the 3' untranslated region. Cell (1993) 74:9–14.[CrossRef][Web of Science][Medline]
  16. Ross J. mRNA stability in mammalian cells. Microbiol Rev (1995) 59:423–450.[Abstract/Free Full Text]
  17. Conne B., Stutz A., Vassalli J.-D. The 3' untranslated region of messenger RNA: A molecular ‘hotspot’ for pathology? Nat Med (2000) 6(6):637–641.[CrossRef][Web of Science][Medline]
  18. Grzybowska E.A., Wilczynska A., Siedlecki J.A. Regulatory functions of 3'UTRs. Biochem Biophys Res Commun (2001) 288:291–295.[CrossRef][Web of Science][Medline]
  19. Kreuzer K.-A., Lass U., Landt O., Nitsche A., Laser J., Ellerbrok H., Pauli G., Huhn D., Schmidt C.A. Highly sensitive and specific fluorescence reverse transcription-PCR assay for the pseudogene-free detection of β-actin transcripts as quantitative reference. Clin Chem (1999) 45(2):297–300.[Free Full Text]
  20. Rasmussen R. Rapid cycle real-time PCR, methods and applications. Meuer S., Wittwer C., Nakagawara K., eds. (2001) Heidenberg: Springer. 21–34.
  21. Pfaffl M.W. A new mathematical model for relative quantification in real-time PCR. Nucleic Acids Res (2001) 29(9):e45.[Abstract/Free Full Text]
  22. Wang X., Kiledjian M., Weiss I.M., Liebhaber S.A. Detection and characterization of a 3' untranslated region ribonucleoprotein complex associated with human {alpha}-globin mRNA stability. Mol Cell Biol (1995) 15:1769–1777.[Abstract]
  23. Searles C.D., Miwa Y., Harrison D.G., Ramasamy S. Posttranscriptional regulation of endothelial nitric oxide synthase during cell growth. Circ Res (1999) 85:588–595.[Abstract/Free Full Text]
  24. Gonzalez-Fernandez F., Jimenez A., Lopez-Blaya A., Velasco S., Arriero M.M., Celdran A., Rico L., Farre J., Casado S., Lopez-Farre A. Cerivastatin prevents tumor necrosis factor-{alpha}-induced downregulation of endothelial nitric oxide synthase: role of endothelial cytosolic proteins. Atherosclerosis (2001) 155:61–70.[CrossRef][Web of Science][Medline]
  25. Tillmar L., Carlsson C., Welsh N. Control of insulin mRNA stability in rat pancreatic islets. Regulatory role of a 3'-untranslated region pyrimidine-rich sequence. J Biol Chem (2002) 277:1099–1106.[Abstract/Free Full Text]
  26. Soderberg M., Raffalli-Mathieu F., Lang M.A. Inflammation modulates the interaction of heterogeneous nuclear ribonucleoprotein (hnRNP) I/polypyrimidine tract binding protein and hnRNP L with the 3'-untranslated region of the murine inducible nitric-oxide synthase mRNA. Mol Pharmacol (2002) 62:423–431.[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
X. Wang, H. Chai, Z. Wang, P. H. Lin, Q. Yao, and C. Chen
Serum amyloid A induces endothelial dysfunction in porcine coronary arteries and human coronary artery endothelial cells
Am J Physiol Heart Circ Physiol, December 1, 2008; 295(6): H2399 - H2408.
[Abstract] [Full Text] [PDF]

Home page
 BloodHome page
C. Chen, H. Chai, X. Wang, J. Jiang, M. S. Jamaluddin, D. Liao, Y. Zhang, H. Wang, U. Bharadwaj, S. Zhang, et al.
Soluble CD40 ligand induces endothelial dysfunction in human and porcine coronary artery endothelial cells
Blood, October 15, 2008; 112(8): 3205 - 3216.
[Abstract] [Full Text] [PDF]

Home page
 Circ. Res.Home page
G. Yan, B. You, S.-P. Chen, J. K. Liao, and J. Sun
Tumor Necrosis Factor-{alpha} Downregulates Endothelial Nitric Oxide Synthase mRNA Stability via Translation Elongation Factor 1-{alpha} 1
Circ. Res., September 12, 2008; 103(6): 591 - 597.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Lung Cell. Mol. Physiol.Home page
V. Brovkovych, X.-P. Gao, E. Ong, S. Brovkovych, M.-L. Brennan, X. Su, S. L. Hazen, A. B. Malik, and R. A. Skidgel
Augmented inducible nitric oxide synthase expression and increased NO production reduce sepsis-induced lung injury and mortality in myeloperoxidase-null mice
Am J Physiol Lung Cell Mol Physiol, July 1, 2008; 295(1): L96 - L103.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Cell Physiol.Home page
C. D. Searles
Transcriptional and posttranscriptional regulation of endothelial nitric oxide synthase expression
Am J Physiol Cell Physiol, November 1, 2006; 291(5): C803 - C816.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Cell Physiol.Home page
M. Baudouin-Legros, A. Hinzpeter, A. Jaulmes, F. Brouillard, B. Costes, P. Fanen, and A. Edelman
Cell-specific posttranscriptional regulation of CFTR gene expression via influence of MAPK cascades on 3'UTR part of transcripts
Am J Physiol Cell Physiol, November 1, 2005; 289(5): C1240 - C1250.
[Abstract] [Full Text] [PDF]

Home page
 Rheumatology (Oxford)Home page
C. D. Buckley, G. E. Rainger, G. B. Nash, and K. Raza
Endothelial cells, fibroblasts and vasculitis
Rheumatology, July 1, 2005; 44(7): 860 - 863.
[Abstract] [Full Text] [PDF]

Home page
 Circ. Res.Home page
C. D. Searles, L. Ide, M. E. Davis, H. Cai, and M. Weber
Actin Cytoskeleton Organization and Posttranscriptional Regulation of Endothelial Nitric Oxide Synthase During Cell Growth
Circ. Res., September 3, 2004; 95(5): 488 - 495.
[Abstract] [Full Text] [PDF]

Home page
 Arterioscler. Thromb. Vasc. Bio.Home page
S. C. Tai, G. B. Robb, and P. A. Marsden
Endothelial Nitric Oxide Synthase: A New Paradigm for Gene Regulation in the Injured Blood Vessel
Arterioscler Thromb Vasc Biol, March 1, 2004; 24(3): 405 - 412.
[Abstract] [Full Text]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Lai, P. F.H
Right arrow Articles by Stewart, D. J
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lai, P. F.H
Right arrow Articles by Stewart, D. J
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?