| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Medizinische Klinik mit Schwerpunkt Kardiologie und Angiologie, Charité – Universitätsmedizin Berlin, CCM, Germany; and
Institut für Biochemie, Justus-Liebig-Universität Giessen, Germany
1Correspondence: Medizinische Klinik mit Schwerpunkt Kardiologie und Angiologie, Charité – Universitätsmedizin Berlin, CCM, Charitéplatz 1, D-10117 Berlin, Germany. E-mail: karl.stangl{at}charite.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: NOS • endothelium • NO • CA repeats
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previously, we have demonstrated an association between high numbers of CA repeats in intron 13 of human eNOS and risk of coronary artery disease (1)
. Interestingly, we have shown that the splicing efficiency of human eNOS can be modulated by the number of CA repeats in intron 13, and we have identified hnRNP L as an activator of eNOS splicing, binding directly to CA repeats (2)
. More importantly, we have recently provided evidence that CA repeats can activate cryptic 5' splice sites in eNOS and regulate 5' splice site choice in other human genes (3)
. Whereas for the other two human NOS proteins iNOS and nNOS, a number of different splice variants including their physiological properties have been described (4)
, no such spliced isoforms are known for eNOS so far.
Alternative splicing plays a major role in control of gene expression and of functions and enzymatic activities of many proteins (5)
. Analysis of pre-mRNA splicing using microarrays indicates that alternatively spliced isoforms exist for more than 70% of all human genes (6)
. Many alternative splice events are restricted to specific tissues or to certain developmental stages (7)
. Recently, evidence has accumulated that alternative splicing is also involved in the onset and progression of human disease (8)
. In Drosophila, a prematurely terminated transcript of the Drosophila NOS (DNOS) led to a hyperproliferative eye phenotype by reducing NOS enzymatic activity and NO production when overexpressed in transgenic flies (9)
.
This study aimed to investigate whether human eNOS is subject to alternative splicing and to determine potential physiological consequences. We have identified shortened alternatively spliced isoforms of human eNOS that contain a novel exon from internal parts of intron 13 and share the same polyadenylation site. The splice variants are differentially expressed in endothelial cells and various human tissues. Expression of one splice variant, eNOS13A, resulted in heterodimerization and reduction of eNOS activity, when coexpressed with full-length eNOS. The expression of eNOS13A was particularly high in testis and selectively up-regulated by long CA repeats in intron 13.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Plasmid construction
The 5'-RACE cDNA clone for eNOS13A in pCRII-TOPO was digested with KpnI and XhoI and inserted into the eukaryotic expression vector pcDNA3.1/V5-His-TOPO (Invitrogen, Carlsbad, CA, USA). To express the tag, an alternative antisense primer, 5'-GAGTCGACTTCTGTGGTGACAGACA-3', in the second round of the 5'-RACE PCR was used to replace the stop codon in eNOS13A by valine. A full-length human eNOS cDNA in pcDNA3.1 was kindly provided by Dr. Dimmeler (10)
.
pcDNAL-CA0, -CA39, and -control: The two genomic clones pB78 and pSac VIS, kindly provided by Dr. Nadaud, were the sources for the eNOS genomic sequences (11)
. A BamHI fragment released from pB78 was cloned into the unique BamHI site of pSac VIS. The new clone, pB78+SacVIS, covers eNOS genomic sequence from intron 9 to intron 15. A PCR fragment amplified using an eNOS exon 13-specific primer EN15, 5'-TTAAATGGTCCTGTGTATGGATGAG-3', and EN16 from intron 13, 5'-TGCCGTTCTAGAGCAGGC-3', was first cloned into pcDNA3.1/V5/His-TOPO vector. EN16 is complementary to an XbaI site in intron 13, which is located 1.4 kb downstream of the 5' splice site of intron 13. A second PCR fragment was amplified using EN1, 5'-TTAAGAATTCGTCCTGTGTATGGATGAGTA-3', and EN2, 5'-TTAATCTAGATTGTGCTGTTCCGGCCGA-3', and digested with XbaI. This XbaI-fragment containing the remaining part of intron 13 and exon 14 was then inserted between the XbaI sites of the new TOPO clone mentioned above. This generated pcDNAL-CA39 carrying 39 CA repeats in intron 13. Plasmids pcDNAL-CA0 and pcDNAL-control were constructed by two-step PCR using pcDNAL-CA39 as template.
Cell culture and transfection
COS-7 cells were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal calf serum (FCS), and HeLa cells in RPMI 1640 medium with 10% FCS. Human umbilical vein endothelial cells (HUVEC) and human aortic endothelial cells (HAEC) were maintained as described previously (12)
. All transfections were done using jetPEI (Biomol, Plymouth Meeting, PA, USA). Plasmids containing full-length eNOS and eNOS13A cDNAs were transfected into COS-7 cells in Opti-MEM medium (Invitrogen) without FCS. In all transfection experiments, the total amount of transfected DNA was kept constant at 10 µg by the addition of empty vector. For proteasome inhibition experiments, 500 nM of the proteasome inhibitor MG132 (Calbiochem, San Diego, CA, USA) or the solvent DMSO in the controls were added to COS-7 cells immediately after transfection and cells were incubated for additional 24 h. To test the effect of CA repeats on eNOS alternative splicing, HeLa cells were transfected with 5 µg of each minigene construct in RPMI 1640 medium containing 10% FCS.
Western blot analysis
After treatment, cells were washed twice with PBS and lysed in extraction buffer (50 mM Tris-HCl pH 7.4, 154 mM KCl, 5 mM glucose, 0.5 mM EDTA, 1x complete protease inhibitor cocktail, and 1% Triton X-100). Total protein (20 µg per lane) was subjected to SDS-PAGE and transferred onto PVDF membranes. After blocking, membranes were incubated with the anti-eNOS antibodies anti-NOSIII 1:2,500 (BD Transduction Laboratories, San Jose, CA, USA), anti-N20 1:2000 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and the anti-V5 antibody 1:5,000 (Invitrogen). After washing, membranes were probed with secondary anti-mouse or anti-rabbit antibodies (Santa Cruz Biotechnology). Bands were visualized by using BCIP or Nitro Blue Tetrazolium (Sigma, St. Louis, MO, USA), or by the ECL detection system (Amersham, Freiburg, Germany).
Low-temperature PAGE
To detect the formation of heterodimers between full-length eNOS and truncated proteins, low-temperature PAGE was performed as described (13)
. Before loading, cell extracts were incubated for 5 min at 95°C for denaturation of proteins or at 37°C for 10 min followed by 30 min at 0°C for preservation of dimerization.
Measurement of eNOS activity
eNOS activity was assessed as described previously (14)
, except that 1 µM FMN and 100 nM calmodulin (CaM) were added to the reaction buffer.
Realtime RT-PCR
Cells were lysed in Trizol reagent (Life Technologies, Inc., Karlsruhe, Germany), and 500 ng of total RNA was reverse-transcribed with random hexamer primers in the case of cultured cells, and with oligo (dT) primers and random hexamers in the case of human tissues. The sequences of primers and TaqMan probes used in this study are provided in the supplementary Table 1. PCR amplifications were performed in a total vol of 25 µl TaqMan Universal Master Mix in a 5700 Sequence Detection System (both Applied Biosystems, Foster City, CA, USA). Thermal cycling conditions comprised an initial denaturation step at 95°C for 10 min, followed by 40 cycles (95°C for 15 s; 60°C for 1 min). Expression levels of the target genes were standardized to ß-actin in HeLa cells and to the geometric mean of GAPDH, HPRT, RPL19, RPS9, and 28S RNA in human tissues and different endothelial cells. For expression analysis in human tissues, the Human Total RNA Master Panel II (Clontech) was used. The expression of the target genes relative to the housekeeping genes was calculated as the difference between the threshold values for the two groups. All PCR reactions were done in duplicates.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Therefore, we searched the database and found an expressed sequence tag (EST) sequence (AW015134) covering part of exon 12, the complete exon 13, and part of intron 13 of human eNOS. To search more systematically for novel eNOS splice isoforms, 5'-RACE was performed with total RNA from HUVEC using primers for exon 1 of human eNOS and a sequence deduced from the AW015134 entry. As a result, three different cDNA clones were recovered, carrying 1800, 2200, and 2300 bp, respectively, of eNOS sequence (Fig. 1
A). The human eNOS gene comprises 26 exons (11)
, the largest intron 13 is 
4.2 kb in length (Fig. 1B
). Complete sequencing of the three cDNA clones revealed sequences identical to exons 1 to 13 of human eNOS for all three clones, followed by novel overlapping exons designated 13A, 13B, and 13C, containing intron 13 sequences (Fig. 1B
). 3'-RACE with specific primers overlapping constitutive exon 13 and alternative exons 13A, 13B, and 13C, respectively, showed that the same polyadenylation signal was used in all three splice variants. The total mRNA lengths for the three splice isoforms are 2030 bp for eNOS13A, 2376 bp for eNOS13B, and 2536 bp for eNOS13C, followed by the poly(A) tail. Alternative exons eNOS13A, B, and C contain 278 bp, 624 bp, and 784 bp, respectively. The cDNA sequences for the three novel eNOS splice variants have been submitted to GenBank with the following accession numbers: DQ256129 for eNOS13A, DQ256130 for eNOS13B, and DQ256131 for eNOS13C. When translated, the alternative exons eNOS13A, B, and C would—after the constitutive exon 13—add 12, 30, and 45 amino acids, respectively. Thus, alternative splicing would result in truncated proteins, lacking the reductase domain of human eNOS and carrying short specific C-terminal sequences.
|
Splice variant eNOS13A suppresses eNOS activity by formation of heterodimers with full-length eNOS
To study potential functional consequences of shortened eNOS transcripts, we chose as an example the splice variant eNOS13A. Transcription and translation of eNOS13A would yield a protein with a predicted size of 65 kDa. The cDNA sequence of eNOS13A was cloned into a mammalian expression vector with a C-terminal V5-tag, and COS-7 cells, which are devoid of eNOS, were cotransfected with a constant amount of full-length eNOS construct (3 µg) and increasing amounts of splice variant eNOS13A (0–7 µg). The overall cDNA quantities were kept constant by the addition of empty vector in all transfections. Figure 2
A gives an overview of the specific antibodies used for the selective detection of both proteins in Western blots. Transfection of eNOS alone produced substantial eNOS activity, whereas eNOS13A alone did not exert eNOS activity above background (Fig. 2B
). Cotransfection of full-length eNOS with increasing amounts of eNOS13A resulted in a concentration-dependent suppression of eNOS activity by up to 40% (Fig. 2B
). Identical results were obtained when the V5-tag was removed, indicating that the C-terminal tag in eNOS13A had no influence on suppression of eNOS activity (data not shown). Vector- and non-transfected COS-7 cells showed only background eNOS activity. Transfection of eNOS13A yielded a protein of the expected size, and the increase in eNOS13A protein levels correlated with the amounts of cDNA transfected (Fig. 2C
, top panel; detection by anti-V5 antibody). Surprisingly, despite a constant amount of full-length eNOS used in all cotransfections, eNOS protein levels decreased, as detected by the anti-NOSIII antibody specific for the C-terminal region (Fig. 2C
, middle panel). The extent of the reduction in eNOS protein levels correlated with the amount of cotransfected eNOS13A. These results were confirmed with the anti-N20 antibody against the N-terminal eNOS region, which detects both proteins (Fig. 2C
, bottom panel).
|
Dimerization is necessary for the enzymatic activity of eNOS (15)
. To investigate whether the truncated splice variant can form heterodimers with full-length eNOS, low-temperature PAGE was performed (Fig. 3
). The splice variant eNOS13A alone formed homodimers, detectable by both the anti-V5 and anti-N20 antibodies, demonstrating that the dimerization domain is preserved in the truncated protein (Fig. 3
, lanes 1 and 2). As expected, full-length eNOS alone gave also homodimers, detectable by both anti-NOSIII and anti-N20 antibodies (Fig. 3
, lanes 3 and 4). Coexpression of full-length eNOS and eNOS13A resulted, besides homodimerization of the two individual proteins, in an additional heterodimer complex (marked by arrows), migrating between the two homodimers formed by either protein (Fig. 3
, lanes 5 and 6). This complex was detectable by all three antibodies, consistent with it representing an eNOS/eNOS13A heterodimer. The heterodimer was most strongly stained with the anti-N20 antibody, which detects both components of the complex, whereas the anti-V5 and anti-NOSIII antibodies detect only one of the two proteins, resulting in weaker staining intensities. In sum, we have demonstrated by coexpression that eNOS13A and full-length eNOS can heterodimerize.
|
The decrease in full-length eNOS protein levels after coexpression with eNOS13 is prevented by inhibition of the proteasome
To test whether the reduction of full-length eNOS protein levels in the presence of coexpressed splice variant eNOS13A are due to enhanced degradation, we inhibited the proteasome. eNOS has been shown to be ubiquitinated and degraded by the proteasome (16)
. COS-7 cells were transfected as above and treated with 500 nM of the proteasome inhibitor MG132 for 24 h; control transfections were done without MG132. Without inhibition of the proteasome, full-length eNOS protein levels decreased in parallel with the overexpression of eNOS13A cDNA, as observed above (Fig. 4
A, panel -MG132; compare with Fig. 2C
). After proteasome inhibition, however, full-length eNOS protein levels did not change significantly, as shown with two independent anti-eNOS antibodies (Fig. 4A
, panel +MG132). The decrease in eNOS activity was comparable with or without inhibition of the proteasome, suggesting that the stabilized heterodimer was not functional (Fig. 4B
).
|
The mRNA levels for full-length eNOS did not decrease below control levels (vector only) after cotransfection with increasing amounts of eNOS13A, suggesting that the regulation of eNOS expression does not occur at the mRNA level (data not shown).
The splice variants are expressed in endothelial cells and various human tissues with eNOS13A showing highest expression in testis
To further characterize the biological significance of these newly discovered eNOS splice variants, we determined whether they were expressed in various endothelial cells and in human tissues (Fig. 5
). First, total RNA was extracted from human umbilical vein endothelial cells (HUVEC), human aortic endothelial cells (HAEC), and non-endothelial COS-7 cells and subjected to realtime RT-PCR with primers specific for each of the splice variants (Fig. 5A
). All three splice variants were expressed in both of the endothelial cells investigated. The relative expression levels of the splice variants in different endothelial cells correlated with full-length eNOS expression. Compared to full-length eNOS, the splice variants were expressed at lower levels (between 0.1% and 5% of full-length eNOS). In COS-7 cells, which do not express eNOS, also none of the splice variants was detectable (data not shown). Second, we determined the expression pattern of the splice variants in various human tissues (Fig. 5B
). In most tissues, all three splice isoforms were detectable. Their relative distribution correlated in general with that of full-length eNOS, with highest expression levels seen in placenta and spleen (marked by arrows). A remarkable exception represents testis, where the expression level of full-length eNOS—as well as that of eNOS13B and eNOS13C—is rather low, whereas expression of eNOS13A is unusually high (marked by an asterisk). The magnitude of eNOS13A expression was around 20% of full-length eNOS. This result was reproduced with testis RNA from a different source (data not shown).
|
Expression of eNOS13A is selectively up-regulated by high CA repeat numbers in intron 13 of human eNOS
Finally, we addressed the question whether the polymorphic CA repeat region in intron 13 of the human eNOS gene might play a role in determining the alternative splicing pattern (Fig. 6
). As shown recently, the CA repeats function as an unusual, length-dependent splice enhancer and can determine 5' splice site choice in human eNOS and other genes (2
, 3)
. To test a potential role of the CA repeat region on the selection of the newly discovered alternative 3' splice sites, we transfected HeLa cells with human eNOS minigene constructs containing the full-length intron 13 with different CA repeat numbers, or with a non-specific sequence as a control, flanked by exons 13 and 14 (Fig. 6A
). Twenty-four hours after transfection, mRNA levels were measured for normally spliced exons 13–14 and for the three splice variants that use alternative 3' splice sites (eNOS13A, B, and C; see Fig. 6B
). Levels of normally spliced minigene and the two splice variants eNOS13B and eNOS13C were not significantly altered, whether intron 13 contained either the control sequence, CA0, or CA39. In contrast, mRNA expression for the eNOS13A splice variant increased approximately 4-fold by intron 13 with 39 CA repeats, as compared with the control and the CA0-containing minigene. In mock-transfected HeLa cells, neither constitutively nor alternatively spliced transcripts were detectable (data not shown). We conclude that high CA repeat numbers in intron 13 of human eNOS selectively stimulate the choice of one of the alternative 3' splice sites, positioned most distally relative to the CA-repeat region (for exon-intron structure, see Fig. 6A
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
NO synthases are a family of proteins—namely inducible (iNOS), neuronal (nNOS) and endothelial (eNOS)—which share a similar genomic organization and exon-intron structure (17)
. Different splice variants are described for nNOS. nNOSµ is tissue-specifically expressed in heart and skeletal muscle of rats and in human penis and urethra (18
, 19)
. Expression of nNOSß was up-regulated in certain brain regions in rats with deletions in exon 2 of nNOS, albeit its catalytic activity was lower compared to full-length nNOS (20)
. Tissue-specific expression of a number of splice variants was also described for human iNOS (21)
, and a potential involvement of iNOS splice variants in B cell viability has been suggested (22)
.
For human eNOS, artificially truncated mutant proteins had been generated to identify the dimerization domain (23)
. Coexpression of these mutant proteins with full-length eNOS in COS-7 cells led to heterodimerization. Both, the N- and C-terminally truncated mutant proteins caused—when coexpressed with wild-type eNOS—a concentration-dependent reduction of eNOS activity, consistent with our results of dominant-negative effects of truncated eNOS splice variants. On the other hand, the authors did not detect a decrease in eNOS activity after addition of purified truncated proteins to endothelial cell extracts, a finding that we confirmed in our study (data not shown) and that indicates that intracellular coexpression of both proteins is necessary for heterodimerization.
There is an interesting parallel to our data in the Drosophila system, where a number of premature terminated transcripts of the Drosophila NOS (DNOS) with enzymatically inactive proteins are expressed during different stages of Drosophila development (24)
. Overexpression of a truncated NOS protein (dNOS4) in transgenic flies inhibited NOS activity by forming complexes with full-length dNOS1 in vivo, and dNOS4 expression suppressed the antiproliferative effects of dNOS1 in the eye and produced a hyperproliferative phenotype in pupae and adult flies (9)
. Similar to NOS, a splice variant of the 
2 subunit of soluble guanylyl cyclase reduced enzyme activity by competing with the full-length protein for heterodimerization (25)
. A truncated isoform of an aminopeptidase II, lacking the C-terminal domain, acted as dominant-negative by formation of heterodimers with full-length protein, and reduced its protein activity (26)
. Taken together, these results point to a general mechanism in regulation of enzyme activity by expression of dominant-negative splice isoforms.
Unlike studies in Drosophila and using artificially truncated proteins of human eNOS (9
, 23
, 24)
, we observed in our study a decrease in full-length eNOS protein levels with increasing amounts of coexpressed eNOS13A cDNA. Non-functional protein complexes are expected to be unstable and hence degraded (27
, 28)
. The proteasome represents the major pathway for protein degradation in eukaryotic cells (29)
, and eNOS has been shown to be ubiquitinated and degraded by the proteasome (16)
. When the proteasome was inhibited, full-length eNOS protein levels remained unchanged even after coexpression of eNOS13A splice variant, suggesting that the proteasome is responsible for degradation. We do not know the basis for this apparent discrepancy between the above studies and our data, but it may be related to differences in the length or stability between the splice variants in either system.
The splice isoforms were all expressed in various endothelial cells and in most human tissues. Particularly high expression of the splice variants as well as normal eNOS mRNA was detected in human placenta and spleen. Interestingly, in human testis one specific splice variant, eNOS13A, was expressed at much higher levels than the other splice variants. This is in agreement with a number of reports that have described peculiar splicing patterns in testis: First, besides the normal full-length eNOS mRNA, a number of smaller transcripts ranging in size from 1.4 to 2.4 kb were detected by Northern blotting specifically in human testis (30)
. Second, a smaller eNOS transcript of 3 kb was observed in rat testis (31)
. For neuronal NOS (nNOS), a transcript with novel 5' exons spliced to exon 4 of full-length nNOS yields an N-terminally truncated protein selectively expressed in human testis (32)
.
How is the expression of the new splice variants regulated? Intron 13 of human eNOS contains polymorphic CA repeats that are genetically determined and varying in length from about CA14 to CA44 (11)
. As we have previously shown, high numbers of CA repeats are associated with an increased risk of coronary artery disease (1)
. Furthermore, in combination with bound heterogenous nuclear RNP (hnRNP) L, intronic CA dinucleotide repeats influence the splicing efficiency of constitutive human eNOS (2)
and can activate alternative 5' splice sites in other human genes (3)
. Transfection of HeLa cells with constructs containing the complete intron 13 with varying lengths of CA repeats resulted in selective up-regulation of eNOS13A at high CA repeat numbers, whereas the relative expression levels of the two other splice variants and of normal eNOS remained unchanged. This suggests that intronic CA repeats, in addition to their known effects on 5' splice sites, can affect alternative 3' splice site selection. That only the expression of eNOS13A was affected by the CA repeat number was a surprising finding, especially considering that in eNOS13A the most distal 3' splice site is activated. Further studies should address the mechanistic basis for this apparent distance requirement.
We found a relatively minor abundance of the splice variants compared to the full-length eNOS protein in endothelial cells and human tissues. However, lower levels of eNOS13 expression were effective in inhibition of eNOS activity in COS-7 cells. Thus, we propose, based on the inhibitory effects on eNOS activity, a potential mechanism for regulating eNOS activity in a moderate long-term manner by the expression of the splice variants in endothelial cells. Modest deregulation of NO production by the dominant-negative effects of the identified splice variants over a long period of time may provide a mechanistic link between the established association between high CA repeat numbers and increased risk of coronary artery disease (1)
. In addition, when simultaneously expressed, the three splice variants may act in an additive manner, resulting in higher impairment of NO production than does a particular splice variant alone.
In summary, we have described here for the first time alternative splice variants for human eNOS that are expressed in endothelial cells and various human tissues. Based on the differential expression of these splice variants and their ability to form heterodimers, we have provided evidence for a potential regulatory mechanism with broad implications in cardiovascular diseases.
|
ACKNOWLEDGMENTS |
|---|
Received for publication October 5, 2006. Accepted for publication December 6, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. Galluccio, P. Piatti, L. Citterio, P. C. G. Lucotti, E. Setola, L. Cassina, M. Oldani, I. Zavaroni, E. Bosi, A. Colombo, et al. Hyperinsulinemia and impaired leptin-adiponectin ratio associate with endothelial nitric oxide synthase polymorphisms in subjects with in-stent restenosis Am J Physiol Endocrinol Metab, May 1, 2008; 294(5): E978 - E986. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
