Regulation of Endothelial Argininosuccinate Synthase Expression and NO Production by an Upstream Open Reading Frame

Argininosuccinate synthase (AS) catalyzes the rate-limiting step in the recycling of citrulline to arginine, which in endothelial cells, is tightly coupled to the production of nitric oxide (NO). In previous work, we established that endothelial AS mRNA can be initiated from multiple start sites, generating co-expressed mRNA variants with different 5′-untranslated regions (5′-UTRs). One of the 5′-UTRs, the shortest form, represents greater than 90% of the total AS mRNA. Two other extended 5′-UTR forms of AS mRNA, resulting from upstream initiations, contain an out-of-frame, upstream open reading frame (uORF). In this study, the function of the extended 5′-UTRs of AS mRNA was investigated. Single base insertions to place the uORF in-frame, and mutations to extend the uORF, demonstrated functionality, both in vitro with AS constructs and in vivo with luciferase constructs. Overexpression of the uORF suppressed endothelial AS protein expression, whereas specific silencing of the uORF AS mRNAs resulted in the coordinate up-regulation of AS protein and NO production. Expression of the full-length of the uORF was necessary to mediate a trans-suppressive effect on endothelial AS expression, demonstrating that the translation product itself affects regulation. In conclusion, the uORF found in the extended, overlapping 5′-UTR AS mRNA species suppresses endothelial AS expression, providing a novel mechanism for regulating endothelial NO production by limiting the availability of arginine. Argininosuccinate synthase (AS) catalyzes the rate-limiting step in the recycling of citrulline to arginine, which in endothelial cells, is tightly coupled to the production of nitric oxide (NO). In previous work, we established that endothelial AS mRNA can be initiated from multiple start sites, generating co-expressed mRNA variants with different 5′-untranslated regions (5′-UTRs). One of the 5′-UTRs, the shortest form, represents greater than 90% of the total AS mRNA. Two other extended 5′-UTR forms of AS mRNA, resulting from upstream initiations, contain an out-of-frame, upstream open reading frame (uORF). In this study, the function of the extended 5′-UTRs of AS mRNA was investigated. Single base insertions to place the uORF in-frame, and mutations to extend the uORF, demonstrated functionality, both in vitro with AS constructs and in vivo with luciferase constructs. Overexpression of the uORF suppressed endothelial AS protein expression, whereas specific silencing of the uORF AS mRNAs resulted in the coordinate up-regulation of AS protein and NO production. Expression of the full-length of the uORF was necessary to mediate a trans-suppressive effect on endothelial AS expression, demonstrating that the translation product itself affects regulation. In conclusion, the uORF found in the extended, overlapping 5′-UTR AS mRNA species suppresses endothelial AS expression, providing a novel mechanism for regulating endothelial NO production by limiting the availability of arginine. IntroductionNitric oxide (NO) 1The abbreviations used are: NO, nitric oxide; AS, argininosuccinate synthase; AL, argininosuccinate lyase; UTR, untranslated region; uORF, upstream open reading frame; nt, nucleotide(s); BAEC, bovine aortic endothelial cells; dnsAUG, downstream AUG; upsStop, upstream stop; GFP, green fluorescent protein; uORFfs, upstream open reading frame shift; siRNA, small interfering RNA. synthesized from arginine by endothelial nitric-oxide synthase is a potent vasodilator and a critical modulator of blood flow and blood pressure. In addition, it mediates vasoprotective actions through inhibiting smooth muscle proliferation, platelet aggregation, and leukocyte adhesion (1Bredt D.S. Snyder S.H. Annu. Rev. Biochem. 1994; 63: 175-195Crossref PubMed Scopus (2127) Google Scholar, 2Gow A.J. Ischiropoulos H. J. Cell. Physiol. 2001; 187: 277-282Crossref PubMed Scopus (130) Google Scholar, 3Vallance P. Chan N. Heart. 2001; 85: 342-350Crossref PubMed Google Scholar). Under pathophysiological conditions associated with endothelial dysfunction, such as heart failure (4Sharma R. Davidoff M.N. Congestive Heart Failure. 2002; 8: 165-172Crossref PubMed Scopus (51) Google Scholar), hypertension, hyper-cholesterolemia, atherosclerosis (5Maxwell A.J. Nitric Oxide. 2002; 6: 101-124Crossref PubMed Scopus (111) Google Scholar), and diabetes (6Goligorsky M.S. Gross S.S. Drug News Perspect. 2001; 14: 133-142PubMed Google Scholar), the ability to produce NO seems to be impaired. Paradoxically, NO production can be impaired by limited availability of the substrate arginine, despite apparently saturating levels of intracellular and extracellular arginine (7Aisaka K. Gross S.S. Griffith O.W. Levi R. Biochem. Biophys. Res. Commun. 1989; 163: 710-717Crossref PubMed Scopus (120) Google Scholar, 8Cooke J.P. Andon N.A. Girerd X.J. Hirsch A.T. Creager M.A. Circulation. 1991; 83: 1057-1062Crossref PubMed Scopus (312) Google Scholar, 9Rossitch Jr., E. Alexander 3rd, E. Black P.M. Cooke J.P. J. Clin. Invest. 1991; 87: 1295-1299Crossref PubMed Scopus (143) Google Scholar, 10Eddahibi S. Adnot S. Carville C. Blouquit Y. Raffestin B. Am. J. Physiol. 1992; 263: L194-L200Crossref PubMed Google Scholar). We have previously shown that under normal conditions, the essential arginine available for NO production is derived from the recycling of citrulline to arginine, catalyzed by two enzymes, argininosuccinate synthase (AS) and argininosuccinate lyase (AL) (11Flam B.R. Hartmann P.J. Harrell-Booth M. Solomonson L.P. Eichler D.C. Nitric Oxide. 2001; 5: 187-197Crossref PubMed Scopus (98) Google Scholar, 12Goodwin B.L. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2004; 279: 18353-18360Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). Although these two enzymes have been studied extensively in liver, where they participate in the urea cycle (13Morris Jr., S.M. Annu. Rev. Nutr. 1992; 12: 81-101Crossref PubMed Scopus (214) Google Scholar), it was not until the discovery of NO that their function in non-hepatic tissues was clarified. In endothelial cells, AS and AL play a critical role in the operation of a citrulline-NO cycle, which supports endothelial NO production (14Sessa W.C. Hecker M. Mitchell J.A. Vane J.R. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 8607-8611Crossref PubMed Scopus (114) Google Scholar, 15Xie L. Gross S.S. J. Biol. Chem. 1997; 272: 16624-16630Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar, 16Xie L. Hattori Y. Tume N. Gross S.S. Semin. Perinatol. 2000; 24: 42-45Crossref PubMed Scopus (31) Google Scholar, 17Su Y. Block E.R. Am. J. Physiol. 1995; 269: L581-L587PubMed Google Scholar).Because AS catalyzes the rate-limiting step in the citrulline-NO cycle (15Xie L. Gross S.S. J. Biol. Chem. 1997; 272: 16624-16630Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar), our initial studies have focused on the molecular basis for the functional role of endothelial AS. Endothelial and hepatic AS appear to have the same primary structure (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar, 19Freytag S.O. Bock H.G. Beaudet A.L. O'Brien W.E. J. Biol. Chem. 1984; 259: 3160-3166Abstract Full Text PDF PubMed Google Scholar), but differ in cellular location and level of expression (11Flam B.R. Hartmann P.J. Harrell-Booth M. Solomonson L.P. Eichler D.C. Nitric Oxide. 2001; 5: 187-197Crossref PubMed Scopus (98) Google Scholar, 18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). Hepatic urea cycle AS and AL are associated with the mitochondria (20Cohen N.S. Kuda A. J. Cell. Biochem. 1996; 60: 334-340Crossref PubMed Scopus (24) Google Scholar), whereas in endothelial cells, AS and AL co-localize with endothelial nitric-oxide synthase in caveolae (11Flam B.R. Hartmann P.J. Harrell-Booth M. Solomonson L.P. Eichler D.C. Nitric Oxide. 2001; 5: 187-197Crossref PubMed Scopus (98) Google Scholar). AS expression in liver also differs from AS expression in endothelial cells as demonstrated by the diversity of co-expressed 5′-UTR AS mRNA species in endothelial cells (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). Three transcription initiation sites identified in endothelial cells result in overlapping 5′-UTR regions of 92, 66, and 43 nucleotides (nt). The longer forms make up ~7% of the total AS message, with the shortest 43-nt 5′-UTR AS mRNA being the predominant species in endothelial cells, and the only detectable form found in liver. Interestingly, the extended 92- and 66-nt 5′-UTR AS mRNAs contain an out-of-frame, upstream overlapping ORF that is terminated by a stop codon 70 nt past the in-frame start codon for the downstream ORF encoding AS. Previously we reported that in vitro translation of AS mRNA containing the extended 5′-UTRs was suppressed compared with the shortest and most predominant 43-nt 5′-UTR AS mRNA species (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). Moreover, we also showed that the translational efficiency of the extended 5′-UTR AS mRNA species was restored to the short form level when the uAUG was mutated to AAG, thus eliminating the uORF (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). This suppression of expression through cis effects was further demonstrated in vivo when the three forms of the AS 5′-UTR were placed in front of a luciferase ORF and transfected into endothelial cells. Here again, the presence of the uAUG found in the extended AS 5′-UTRs suppressed expression of luciferase in a cis-dependent manner.Upstream ORFs can affect the translation of a downstream ORF in a variety of ways (21Morris D.R. Geballe A.P. Mol. Cell. Biol. 2000; 20: 8635-8642Crossref PubMed Scopus (560) Google Scholar). In higher eukaryotes, initiation of translation generally occurs at the first AUG that resides in a favorable context. When the first AUG context is suboptimal, a portion of the scanning ribosomes may continue past the first AUG and initiate translation downstream at subsequent AUGs via leaky scanning (22Kozak M. Gene (Amst.). 1999; 234: 187-208Crossref PubMed Scopus (1121) Google Scholar). Several eukaryotic mRNAs have been shown to contain one or more ORFs that affect the translational efficiency of the main, downstream ORF (21Morris D.R. Geballe A.P. Mol. Cell. Biol. 2000; 20: 8635-8642Crossref PubMed Scopus (560) Google Scholar). Depending on factors such as intercistronic length and secondary structure, scanning ribosomes, upon initiation at the uAUG, can either translate the uORF and reinitiate downstream or stall on the mRNA during elongation, thus preventing initiation at other sites (21Morris D.R. Geballe A.P. Mol. Cell. Biol. 2000; 20: 8635-8642Crossref PubMed Scopus (560) Google Scholar). In other cases, partial translation of the nascent peptide prevents downstream re-initiation by interaction of the peptide with a protein or RNA in the ribosome preventing termination from proceeding efficiently (23Gaba A. Wang Z. Krishnamoorthy T. Hinnebusch A.G. Sachs M.S. EMBO J. 2001; 20: 6453-6463Crossref PubMed Scopus (114) Google Scholar). However, another less common event is for the uORF to be translated and for the peptide product to affect translation of the downstream cistron via a trans mechanism (24Parola A.L. Kobilka B.K. J. Biol. Chem. 1994; 269: 4497-4505Abstract Full Text PDF PubMed Google Scholar). Based on these examples and our previous findings, we show in this report that the uORF in the extended 5′-UTR AS mRNA species is functional and acts to limit overall AS expression as well as NO production, thus providing a novel mechanism for regulating endothelial NO production.DISCUSSIONWe previously established that the recycling of citrulline to arginine is essential to provide the substrate arginine for NO production, even in the presence of saturating levels of intra- and extracellular arginine (11Flam B.R. Hartmann P.J. Harrell-Booth M. Solomonson L.P. Eichler D.C. Nitric Oxide. 2001; 5: 187-197Crossref PubMed Scopus (98) Google Scholar, 12Goodwin B.L. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2004; 279: 18353-18360Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). We demonstrate in this study that expression of the extended 5′-UTR forms of AS mRNA, containing an uORF, mediates a trans effect, suppressing overall endothelial AS expression and causing a corresponding suppression of endothelial NO production. This suppression of AS expression requires a functional, out-of-frame uORF represented in the 5′-UTR regions of the co-expressed extended forms of endothelial AS mRNA (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). The uORF AUG was shown to be functional both in vitro and in vivo. When the uAUG was put in-frame with the downstream AUG by inserting a nucleotide, two in vitro translated 35S-labeled products were evidenced by electrophoretic SDS-polyacrylamide gel analysis. The larger AS species (~49 kDa) was initiated from the uAUG, whereas the smaller (~47 kDa) species represented the translation product initiated from the normal, downstream reading frame encoding AS. Interestingly, the ratio of products in this case favored use of the uAUG. Moreover, when the context of this uAUG was altered to better match the Kozak consensus initiation sequence (22Kozak M. Gene (Amst.). 1999; 234: 187-208Crossref PubMed Scopus (1121) Google Scholar), translation significantly improved from the uAUG. To demonstrate that this uORF, when positioned out-of-frame, was still translated, two putative stop codons for the uORF were mutated to allow production of a larger, more easily identifiable translation product (~21 kDa). Although the difference in methionine content did not permit a quantitative comparison by 35S labeling, the results clearly demonstrated a 21-kDa product, confirming the functionality of the uORF in its natural context.With the support of in vitro results, we then assessed the in vivo functionality of the uORF in endothelial cells using a luciferase reporter assay. Expression of luciferase from the uAUG demonstrated that the context of the uAUG is sufficient to support initiation of translation. Moreover, when the AS uAUG start codon was positioned in-frame, in the context of the entire 5′-UTR and preceding the normal start codon for a luciferase gene, our results demonstrated In this two luciferase products were identified by with the that both the uAUG and the downstream luciferase AUG are in endothelial from our that AS mRNA species containing the uORF in the extended 5′-UTR sequence not AS either in vitro or in of cis effects of the uORF (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). In this we have not only the functionality of the uORF, but also its that of this uORF resulted in a in AS expression in endothelial cells. This result that the of the extended 5′-UTR forms of AS mRNA, containing an out-of-frame uORF, may play a role in suppressing the overall expression of endothelial AS. we showed that NO production is significantly when the AS uORF is further the for AS expression to NO production in endothelial cells. The that AS expression was not suppressed when the uORF was via of an start or by of containing either a or altered start or stop demonstrated that the entire sequence of the uORF is to mediate the trans effects that endothelial AS expression and NO production. a effect was that expression of the translational product by the uORF to the suppression of endothelial AS expression of the endothelial extended 5′-UTR AS mRNA species were by expression of AS despite the that these species less than of the total AS mRNA. with the rate-limiting role of AS in recycling citrulline to arginine and in the essential arginine for NO production, of the extended 5′-UTR AS mRNA species containing this uORF resulted in an of endothelial cells to produce the overall results that the uORF found in the extended 5′-UTR forms of endothelial AS mRNA is functional, and as such a protein product that acts to expression of the predominant short form of the AS a small protein through expression of the uORF of the extended 5′-UTRs of two forms of AS mRNA, to endothelial cells, suppresses AS The overall effect of this suppression of AS expression is to NO production in endothelial cells by limiting the availability of the substrate arginine. results provide for a novel mechanism for the of endothelial AS protein expression and further support the essential role of the citrulline-NO cycle in endothelial NO production. IntroductionNitric oxide (NO) 1The abbreviations used are: NO, nitric oxide; AS, argininosuccinate synthase; AL, argininosuccinate lyase; UTR, untranslated region; uORF, upstream open reading frame; nt, nucleotide(s); BAEC, bovine aortic endothelial cells; dnsAUG, downstream AUG; upsStop, upstream stop; GFP, green fluorescent protein; uORFfs, upstream open reading frame shift; siRNA, small interfering RNA. synthesized from arginine by endothelial nitric-oxide synthase is a potent vasodilator and a critical modulator of blood flow and blood pressure. In addition, it mediates vasoprotective actions through inhibiting smooth muscle proliferation, platelet aggregation, and leukocyte adhesion (1Bredt D.S. Snyder S.H. Annu. Rev. Biochem. 1994; 63: 175-195Crossref PubMed Scopus (2127) Google Scholar, 2Gow A.J. Ischiropoulos H. J. Cell. Physiol. 2001; 187: 277-282Crossref PubMed Scopus (130) Google Scholar, 3Vallance P. Chan N. Heart. 2001; 85: 342-350Crossref PubMed Google Scholar). Under pathophysiological conditions associated with endothelial dysfunction, such as heart failure (4Sharma R. Davidoff M.N. Congestive Heart Failure. 2002; 8: 165-172Crossref PubMed Scopus (51) Google Scholar), hypertension, hyper-cholesterolemia, atherosclerosis (5Maxwell A.J. Nitric Oxide. 2002; 6: 101-124Crossref PubMed Scopus (111) Google Scholar), and diabetes (6Goligorsky M.S. Gross S.S. Drug News Perspect. 2001; 14: 133-142PubMed Google Scholar), the ability to produce NO seems to be impaired. Paradoxically, NO production can be impaired by limited availability of the substrate arginine, despite apparently saturating levels of intracellular and extracellular arginine (7Aisaka K. Gross S.S. Griffith O.W. Levi R. Biochem. Biophys. Res. Commun. 1989; 163: 710-717Crossref PubMed Scopus (120) Google Scholar, 8Cooke J.P. Andon N.A. Girerd X.J. Hirsch A.T. Creager M.A. Circulation. 1991; 83: 1057-1062Crossref PubMed Scopus (312) Google Scholar, 9Rossitch Jr., E. Alexander 3rd, E. Black P.M. Cooke J.P. J. Clin. Invest. 1991; 87: 1295-1299Crossref PubMed Scopus (143) Google Scholar, 10Eddahibi S. Adnot S. Carville C. Blouquit Y. Raffestin B. Am. J. Physiol. 1992; 263: L194-L200Crossref PubMed Google Scholar). We have previously shown that under normal conditions, the essential arginine available for NO production is derived from the recycling of citrulline to arginine, catalyzed by two enzymes, argininosuccinate synthase (AS) and argininosuccinate lyase (AL) (11Flam B.R. Hartmann P.J. Harrell-Booth M. Solomonson L.P. Eichler D.C. Nitric Oxide. 2001; 5: 187-197Crossref PubMed Scopus (98) Google Scholar, 12Goodwin B.L. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2004; 279: 18353-18360Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). Although these two enzymes have been studied extensively in liver, where they participate in the urea cycle (13Morris Jr., S.M. Annu. Rev. Nutr. 1992; 12: 81-101Crossref PubMed Scopus (214) Google Scholar), it was not until the discovery of NO that their function in non-hepatic tissues was clarified. In endothelial cells, AS and AL play a critical role in the operation of a citrulline-NO cycle, which supports endothelial NO production (14Sessa W.C. Hecker M. Mitchell J.A. Vane J.R. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 8607-8611Crossref PubMed Scopus (114) Google Scholar, 15Xie L. Gross S.S. J. Biol. Chem. 1997; 272: 16624-16630Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar, 16Xie L. Hattori Y. Tume N. Gross S.S. Semin. Perinatol. 2000; 24: 42-45Crossref PubMed Scopus (31) Google Scholar, 17Su Y. Block E.R. Am. J. Physiol. 1995; 269: L581-L587PubMed Google Scholar).Because AS catalyzes the rate-limiting step in the citrulline-NO cycle (15Xie L. Gross S.S. J. Biol. Chem. 1997; 272: 16624-16630Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar), our initial studies have focused on the molecular basis for the functional role of endothelial AS. Endothelial and hepatic AS appear to have the same primary structure (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar, 19Freytag S.O. Bock H.G. Beaudet A.L. O'Brien W.E. J. Biol. Chem. 1984; 259: 3160-3166Abstract Full Text PDF PubMed Google Scholar), but differ in cellular location and level of expression (11Flam B.R. Hartmann P.J. Harrell-Booth M. Solomonson L.P. Eichler D.C. Nitric Oxide. 2001; 5: 187-197Crossref PubMed Scopus (98) Google Scholar, 18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). Hepatic urea cycle AS and AL are associated with the mitochondria (20Cohen N.S. Kuda A. J. Cell. Biochem. 1996; 60: 334-340Crossref PubMed Scopus (24) Google Scholar), whereas in endothelial cells, AS and AL co-localize with endothelial nitric-oxide synthase in caveolae (11Flam B.R. Hartmann P.J. Harrell-Booth M. Solomonson L.P. Eichler D.C. Nitric Oxide. 2001; 5: 187-197Crossref PubMed Scopus (98) Google Scholar). AS expression in liver also differs from AS expression in endothelial cells as demonstrated by the diversity of co-expressed 5′-UTR AS mRNA species in endothelial cells (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). Three transcription initiation sites identified in endothelial cells result in overlapping 5′-UTR regions of 92, 66, and 43 nucleotides (nt). The longer forms make up ~7% of the total AS message, with the shortest 43-nt 5′-UTR AS mRNA being the predominant species in endothelial cells, and the only detectable form found in liver. Interestingly, the extended 92- and 66-nt 5′-UTR AS mRNAs contain an out-of-frame, upstream overlapping ORF that is terminated by a stop codon 70 nt past the in-frame start codon for the downstream ORF encoding AS. Previously we reported that in vitro translation of AS mRNA containing the extended 5′-UTRs was suppressed compared with the shortest and most predominant 43-nt 5′-UTR AS mRNA species (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). Moreover, we also showed that the translational efficiency of the extended 5′-UTR AS mRNA species was restored to the short form level when the uAUG was mutated to AAG, thus eliminating the uORF (18Pendleton L.C. Goodwin B.L. Flam B.R. Solomonson L.P. Eichler D.C. J. Biol. Chem. 2002; 277: 25363-25369Abstract Full Text Full Text PDF PubMed Scopus (19) Google Scholar). This suppression of expression through cis effects was further demonstrated in vivo when the three forms of the AS 5′-UTR were placed in front of a luciferase ORF and transfected into endothelial cells. Here again, the presence of the uAUG found in the extended AS 5′-UTRs suppressed expression of luciferase in a cis-dependent manner.Upstream ORFs can affect the translation of a downstream ORF in a variety of ways (21Morris D.R. Geballe A.P. Mol. Cell. Biol. 2000; 20: 8635-8642Crossref PubMed Scopus (560) Google Scholar). In higher eukaryotes, initiation of translation generally occurs at the first AUG that resides in a favorable context. When the first AUG context is suboptimal, a portion of the scanning ribosomes may continue past the first AUG and initiate translation downstream at subsequent AUGs via leaky scanning (22Kozak M. Gene (Amst.). 1999; 234: 187-208Crossref PubMed Scopus (1121) Google Scholar). Several eukaryotic mRNAs have been shown to contain one or more ORFs that affect the translational efficiency of the main, downstream ORF (21Morris D.R. Geballe A.P. Mol. Cell. Biol. 2000; 20: 8635-8642Crossref PubMed Scopus (560) Google Scholar). Depending on factors such as intercistronic length and secondary structure, scanning ribosomes, upon initiation at the uAUG, can either translate the uORF and reinitiate downstream or stall on the mRNA during elongation, thus preventing initiation at other sites (21Morris D.R. Geballe A.P. Mol. Cell. Biol. 2000; 20: 8635-8642Crossref PubMed Scopus (560) Google Scholar). In other cases, partial translation of the nascent peptide prevents downstream re-initiation by interaction of the peptide with a protein or RNA in the ribosome preventing termination from proceeding efficiently (23Gaba A. Wang Z. Krishnamoorthy T. Hinnebusch A.G. Sachs M.S. EMBO J. 2001; 20: 6453-6463Crossref PubMed Scopus (114) Google Scholar). However, another less common event is for the uORF to be translated and for the peptide product to affect translation of the downstream cistron via a trans mechanism (24Parola A.L. Kobilka B.K. J. Biol. Chem. 1994; 269: 4497-4505Abstract Full Text PDF PubMed Google Scholar). Based on these examples and our previous findings, we show in this report that the uORF in the extended 5′-UTR AS mRNA species is functional and acts to limit overall AS expression as well as NO production, thus providing a novel mechanism for regulating endothelial NO production.