OUP user menu

Molecular control of nitric oxide synthases in the cardiovascular system

Andreas Papapetropoulos , Radu Daniel Rudic , William C Sessa
DOI: http://dx.doi.org/10.1016/S0008-6363(99)00161-3 509-520 First published online: 15 August 1999


Nitric oxide plays an important role in cardiovascular homeostasis. In this review, the regulation of the three nitric oxide synthase isoforms in the cardiovascular system are examined at molecular and cellular levels. In addition, recent information gleaned from the use of NOS knockout mice are discussed.

  • Nitric oxide synthases
  • Molecular control
  • Cardiovascular system

Time for primary review 33 days.

1 Introduction

Nitric oxide (NO) is a pluripotent regulatory gas in the cardiovascular system generated by the nitric oxide synthase (NOS) family of proteins. NOSs are NAPDH-requiring heme containing oxido-reductases that convert one of the chemically equivalent guanidino nitrogens of l-arginine to NO and the by product l-citrulline. At least three distinct NOS isoforms exist in mammalian cells: neuronal (nNOS; Type I), inducible (iNOS, Type II) and endothelial (eNOS or Type III). All NOSs have approximately 60% amino acid identity and very similar primary structures. The purpose of this review is to summarize recent developments in the molecular biology and biochemistry of NOS with a special emphasis on the functional roles for NO in the cardiovascular system gleaned from genetic approaches.

2 Regulation of NOS expression:

2.1 nNOS

nNOS was the first of the three isoforms to be purified and cloned [1]. Although its name implies restricted expression to neuronal tissues, it is now well established that nNOS is present outside the central and peripheral nervous system and may be involved in several aspects of cardiovascular control. Recently, it has been shown that NO produced by perivascular nerves of cerebral arteries directly modulates vascular tone [2]. In vascular cells, expression of nNOS mRNA transcripts and/or immunoreactivity, has been shown to occur in endothelial [3] as well as smooth muscle cells of rat and human origin [4,5]. As is a common theme for all NOS isoforms, the mRNA and protein is found in cells types where the functional role of NO is yet to be discovered. Recently, the presence of catalytically active nNOS was shown in vascular smooth muscle cells of rat carotid arteries. NO production from endothelium-denuded carotid arteries (as evidenced using a NOS inhibitor) was enhanced in old, but not young spontaneously hypertensive rats (SHR) or WKY controls, suggesting that persistent hypertension unmasks the functional importance of nNOS, ex vivo [4]. The mechanism for this is unclear, as both control and young SHR express nNOS in smooth muscle. Possibly, the gain of nNOS function seen in old SHR may be due to post-translational control mechanisms, such as the loss of a protein inhibitor of nNOS activity, namely PIN [6], CAPON [7] or caveolin [8,9]. Additionally, the increased production of NO from nNOS has been shown to mediate dilation of pial arteries in eNOS knockout mice, but not in littermate controls [10]. However, in this study it is unclear if the increased nNOS expression occurs in vascular cells or in perivascular nerves. Finally, there is evidence in human atherosclerotic lesions that nNOS is expressed in endothelial and mesenchymal-like intimal cells [11].

2.2 Transcriptional regulation

The human nNOS gene occupies approximately 200 kb of DNA and its major neuronal transcript consists of 29 exons and 28 introns [12]. The 5’flanking region of this gene reveals the presence of multiple potential cis-acting elements such as AP-2, TEF-1/MCBF, CREB/ATF/c-fos, NRF-1, Ets, NF-1 and NF-kB binding motifs. However, its still unclear which of these elements in the promoter/enhancer region, if any, contribute to the transcriptional regulation of nNOS expression. Although nNOS is considered a ‘constitutively’ expressed gene, upregulation of nNOS mRNA and protein are thought to be a generalized response of neuronal cells to biological, physical or chemical stress-inducing stimuli such as spinal cord and nerve injuries, immobilization stress, electrical stimulation, and colchicine or formalin administration. Enhanced nNOS levels coincide temporally and spatially with increased levels of the transcription factors c-fos and c-jun. No studies investigating the regulation of nNOS expression in the cardiovascular system has been reported. The reader is referred to recent reviews for further information on the control of nNOS expression in neuronal tissues [13,14].

2.3 Transcript diversity–alternative splicing.

Molecular diversity of the nNOS gene is manifested by the presence of multiple transcripts. These multiple transcripts arise from the use of multiple promoters, exon deletions or insertions and the use of alternate polyadenylation signals. It is now well accepted that two major transcriptional clusters exist within the human nNOS gene, the neuronal and the testis-specific cluster. The neuronal cluster is found upstream of exon 2. At least eight different first exons are spliced to a common second exon. Since the translational start site is found within exon 2, all of these transcripts should give rise to the same protein [15,16]. The testis-specific transcript originates from within intron 3, giving rise to an mRNA species with a 5′ terminus encoded for by two new exons spliced to exon 4 of the full-length nNOS [17]. The testis-specific nNOS translates to a smaller protein with calculated molecular mass of 125 kDa that is expressed at comparable levels and has similar catalytic activity to full-length nNOS. The advantage of such complexity in the nNOS transcripts is not obvious, but it may facilitate tissue specific and/or developmental regulation of nNOS expression. Since the size of nNOS protein expressed in vascular cells has been reported to be approximately 150 kDa, it is unlikely that the testis-specific form of nNOS is expressed in these cells. Moreover, the expression of another nNOS transcript, nNOSμ, has been confirmed to occur in skeletal, but not smooth muscle or in cardiomyocytes [18]. nNOSμ contains an extra 102 bp between exons 16 and 17 and in spite of carrying 34 additional amino acids at the carboxy side of the calmodulin binding domain it has similar catalytic activity with that expressed in the rat cerebellum and correlates with myotube fusion.

In addition to the differences in the 5′ UTR of nNOS, cassette deletions of exons 9/10 and 10 have been demonstrated in human and murine tissues and human cell lines [19]. Alternative splicing results in an mRNA species that is 315 bp shorter (exons 9/10). This form is expressed at lower levels in many areas of the nervous system and gives rise to a protein lacking 105 amino acids located at the amino terminal of the calmodulin-binding domain. This shorter nNOS protein retains NADPH-diaphorase activity while it is unable to convert arginine to citrulline. Splice variants lacking exon 10 have also been described, but are expected to yield an inactive protein since a premature stop codon is introduced. Whether these alternatively spliced nNOS mRNAs are translated in vivo and can act as ‘endogenous dominant negative’ alleles remains to be elucidated. Existence of these splice variants has not been confirmed in cardiovascular tissue.

2.4 Post-translational regulation

nNOS has been shown to interact with several proteins that either determine targeting or alter its activity including members of the dystrophin family of proteins [20], post-synaptic density proteins 93 and 95 [21], PIN [6], CAPON [7] and caveolin-3 [9] in skeletal muscle, neuronal cells or transfected systems. Recently, we have demonstrated that nNOS and caveolin-3 co-distribute in arterial, but not venous vascular smooth muscle, suggesting differential control of nNOS function in these vascular beds by caveolin-3 [22].

nNOS has also been shown to be phosphorylated by a variety of serine kinases including protein kinase A, calmodulin-dependent kinases, and protein kinase C [23,24]. nNOS in astrocytoma cells is also phosphorylated on tyrosine residues in response to lipopolysaccharide and interferon, suggesting a link between signal transduction and NOS activation [25]. However, the sites of nNOS phosphorylation and the functional significance of this modification in neural or vascular cells are not yet known.

2.5 iNOS

Unlike nNOS and eNOS, expression of iNOS has been documented in all nucleated cells in the cardiovascular system, including most vascular endothelial and endocardial cells, vascular smooth muscle and cardiac myocytes, as well as in inflammatory cells found in the subendothelial space, such as leukocytes, fibroblasts and mast cells. iNOS expression is transcriptionally regulated following cytokine (tumor necrosis factor-α; TNF-α, interleukin-1β; IL-1β, IL-2 or interferon-γ; IFN-γ) or bacterial lipopolysaccharide (LPS) stimulation. In general, cells derived from rodents are more easily stimulated to produce NO from iNOS than are human cells, at least in vitro [26,27].

2.6 Transcriptional regulation

Most of the work on the transcriptional regulation of iNOS expression has been performed using freshly isolated murine macrophages, macrophage cell lines or cultured rat aortic vascular smooth muscle cells (VSM). Expression of iNOS in these cells mediates their cytotoxic actions and plays an important role in the immune response. Cloning of a 1.7 kb fragment flanking the transcriptional start site of the murine gene reveals several putative transcription factor binding sequences including ten IFN-γ response elements (IFN-RE), three γ-activated sites (GAS), two consensus sequences for nuclear factor-kB (NF-kB) binding and four for NF-IL6, two TNF-α response elements (TNFα-RE), two activating protein-1 binding motifs (AP-1), three interferon-α stimulated response elements (ISRE), and a basal transcription recognition site (TATA box) [28,29]. Many of these elements are also present in the human iNOS promoter [30].

In spite of the presence of several putative cis-acting regulatory domains in the iNOS promoter, the importance of only two of them has been documented to be important for iNOS induction. The NF-kB site within -76 to -85 of the murine promoter is essential for LPS-induced iNOS transcription, with p50/c-rel and p50/RelA heterodimers identified as components of this transacting factor [31]. A different series of experiments, confirmed that a cluster of four enhancer elements known to bind IFN-γ responsive transcription factors (-911 to -951) are important for iNOS transcription. Mutagenic analysis and gel shifts revealed that IRF-1 is responsible for this regulation [32]. The functional confirmation for the importance of IRF-1 driving iNOS expression was obtained from mice with targeted disruption of the IRF-1 gene. Macrophages isolated from IRF-1 knockout mice fail to respond to stimulation by increasing iNOS mRNA and nitrite production [33].

Control of iNOS expression in the rat VSM cell line A7r5 has been studied [34]. A 1.7 kb 5′ upstream fragment of the murine iNOS gene functions as a promoter within these cells but not as efficiently as in murine macrophages. A major difference in the behavior of the iNOS promoter between VSM and murine macrophages was shown using promoter constructs that carried mutations in one or both NF-kB sites with the upstream site (-962 to -971) being most important for induction in VSM and the downstream site (-76 to -85) important in murine macrophages. Deletion of both sites practically abolished induction by the cytokine mixture in VSM. However, it should be noted that the VSM cell line used was unresponsive to LPS stimulation. In addition, the protein binding to the necessary NF-kB site was identified as p65 together with an unidentified 50kDa protein in VSM.

2.7 Post-transcriptional regulation

Regulation of iNOS expression at the post-transcriptional level has also been described. The 3′ untranslated region of the iNOS mRNA contains a ‘AUUUA’ motif which can potentially destabilize mRNAs. This is reflected by the short half-life of the iNOS mRNA in freshly isolated murine macrophages (approximately 3 h); however direct demonstration that this motif directs mRNA destabilization is lacking. Potentiation of IFN-γ induced iNOS gene expression in RAW 264.7 cells by LPS is due to an increase the mRNA half-life from 1.5 h in cells stimulated with IFN-γ to 6 h in cells treated with both LPS and IFN-γ. This observation provides a potential molecular mechanism for the synergistic effects of LPS with IFN-γ. The potentiating actions of IFN on iNOS mRNA levels is similarly seen with cAMP and agents that increase this second messenger in cardiac myocytes.

In contrast to the stabilizing action of LPS on iNOS mRNA levels in IFN-γ stimulated macrophages, TGF-β was shown to decrease iNOS expression by decreasing mRNA stability. In addition, TGF-β decreases translation of mRNA without affecting the rate of transcription, and also increases iNOS protein degradation and activity [35]. Postranscriptional regulation of iNOS expression has also been demonstrated in rat VSM cells as treatment with cycloheximide causes a superinduction of iNOS mRNA levels by prolonging its half-life from 2–3 h to greater than 12 h [36]. However, extrapolation of these results in vivo are not clear since neither the synergy between LPS and IFN-γ, nor the superinduction by cycloheximide, was found in rat aortic strips [37]. TGF-β, inhibits iNOS induction in VSM similar to its actions in murine macrophages. However, the only proven action of TGF-β in VSM is through a reduction in iNOS transcription, a mechanism different from that observed in the murine macrophages [38].

2.8 Post-translational regulation

Most of the efforts to understand the regulation of this NOS isoform have focused on the events leading to transcriptional activation of the iNOS gene. To date, there is no published data demonstrating that iNOS can interact with other cellular proteins besides calmodulin [39,40] and very little on understanding its subcellular trafficking [41,42]. iNOS is phosphorylated in macrophages and this may occur on serine and tyrosine [42,43]. The specific sites of phosphorylation and its importance in regulation of iNOS activity are not known.

2.9 eNOS

eNOS protein was originally purified and the corresponding cDNA cloned from endothelial cells (EC) and is the NOS isoform responsible for producing the classic endothelium-derived relaxing factor as described by Furchgott [44–47]. Further experimentation has revealed that more cells within the cardiovascular system express eNOS including cardiac myocytes [48] and platelets [49]. In myocytes, locally produced NO may modulate inotropic and chronotropic state of the heart as NOS inhibitors influence the force of contraction and spontaneous rate of beating in isolated myocytes. Since little in general in known about the molecular mechanisms of eNOS expression in different cell types, the regulation of eNOS will be discussed in the context of EC.

2.10 Transcriptional and post-transcriptional regulation

Cloning of the human and bovine eNOS genes revealed a 5′ regulatory region containing a ‘TATA-less’ promoter and a variety of cis elements for the putative binding of transcription factors [50,51]. Specific sites that may influence basal transcription found in the eNOS gene include Sp1 and GATA sites, whereas potential sites for stimulated transcription include a sterol regulatory element (SRE), activator proteins 1 and 2 elements (AP1, AP2), a nuclear factor-1 element (NF-1), partial estrogen responsive elements ERE), a cAMP response element (CRE) and a putative shear stress response element (SSRE).

Initial experiments using reporter gene constructs of the human eNOS promoter reveals that the proximal Sp1 is necessary for basal transcription [52,53]. Deletion or site directed mutagenesis of this site reduces promoter activity by 90–95% when transfected into EC. In addition to Sp1 regulating basal transcription, GATA-2 a transcription factor highly expressed in EC, also modulates basal activity. Recently, the expression of the human eNOS promoter in transgenic mice demonstrates that the promoter can direct the expression of the transgene into vascular endothelial cells of conduit vessels and large and small blood vessels of the heart, brain and skeletal muscle, but not to the vasculature of lung or liver [54]. This suggests that additional regulatory elements are required for uniform expression in all endothelia and the 1600 bp fragment contains information for regional expression. Several pharmacological agents or physiological situations have been reported to alter eNOS gene expression, such as shear stress or exercise training [55–57] however, the molecular mechanisms by which they occur are virtually unknown. In the present review, only agents for which the molecular mechanism has been elucidated in some detail are going to be discussed, the reader is referred to recent reviews for a more complete listing of agents altering eNOS expression (see [19].

The most potent activator of eNOS expression is lysophosphotidylcholine (LPC). LPC content of atherosclerotic blood vessels is higher than that of normal vessels and is a major phospholipid found in oxidized low density lipoproteins. Treatment of human umbilical vein EC with LPC initiates a dose- and time-dependent induction of eNOS mRNA [58]. The magnitude of the mRNA induction is 11-fold with a consistent, but not equal, increase in the transcription rate of the eNOS gene, as measured in nuclear run-off studies. NOS protein levels, activity and biologically active NO are all increased with LPC treatment. However, there is a clear discrepancy between the levels of eNOS mRNA levels, protein and activity suggesting complex post-transcriptional and post-translational regulation. These data are consistent with higher levels of eNOS protein found in atherosclerotic rabbits [59]. The stimulatory effects of LPC on eNOS gene expression have been reproduced in other species of EC, albeit to lesser extent [60]. A recent study has demonstrated that the transcription factor Sp1 is necessary for the actions of LPC, since LPC increases the binding of Sp1 to the eNOS promoter. Moreover, LPC activates the nuclear serine phosphatase (PP2A) which is involved with the ability of LPC to induce eNOS expression [61].

In contrast to the stimulatory effects of LPC on eNOS, Liao et al. reported that oxidized LDL decreases eNOS expression in human EC [62]. At early time points, the suppression appears to be mediated through a partial decrease in the rate of transcription and a destabilization of the NOS mRNA whereas at later time points, the rate of transcription increases paradoxically relative to the reduction in steady state mRNA levels. However, the relevance of greater eNOS expression in atherosclerosis is unknown. Perhaps, during development of atherosclerosis prior to lesion formation, LPC induction of eNOS affords protection against pro-atherosclerotic mechanisms (recruitment of mononuclear cells and oxidation of low density lipoproteins, LDL) required for the lesion development. The ability of lipids derived from LDL and oxidized LDL, per se, to influence eNOS expression and activity clearly merits further evaluation.

A recent discovery germane to the importance of eNOS in atherosclerosis are the findings that statin based cholesterol lowering drugs modestly increased eNOS mRNA levels by a post-transcriptional mechanism involving eNOS mRNA stabilization [63]. The mechanisms for these effects are not due to the lowering of intracellular cholesterol levels, per se, but are likely mediated via inhibition of the lipidation of the small GTP binding protein, Rho [64]. Indeed, mevastatin inhibits the geranyl-geranylation and subsequent translocation of Rho from the cytosol to the membrane, an essential step for Rho activation, concomitantly with an induction of eNOS mRNA levels. Furthermore, stimulation of Rho GTPase decreases eNOS expression while inhibition of Rho enhances eNOS mRNA and protein levels. The precise link between Rho and eNOS mRNA stability is not clear but may relate to the role of Rho or downstream effectors of Rho, in modulation of the actin cytoskeleton. Most importantly, the induction of eNOS by statins is functionally relevant as HMG-CoA inhibitors reduce stroke in mice, an effect completely absent in eNOS knockout mice [65], underscoring the fundamental importance of this discovery.

Additionally TGF-β, tumor necrosis factor-α (TNFα) and estrogen influence eNOS mRNA levels. TGF-β induces eNOS mRNA, protein and activity in bovine aortic EC [66]. Transient transfection assays using bovine eNOS promoter–reporter constructs demonstrates that the TGF-β responsive element resides between -935 and -1269 nt upstream of the transcriptional start site. Gel shift assays and point mutation analysis show that TGF-β increases the binding of the CCAAT transcription factor/nuclear factor-1 to the NF-1 site (-1014 to -1026 nt). However, this site is necessary but not sufficient for TGF-β activation of eNOS suggesting that additional factors are most likely required. TNFα down-regulates eNOS activity, protein and mRNA in endothelial cells [67]. This downregulation has been attributed to destabilization of mRNA that results from interaction of a TNFα inducible protein that binds to the 3′ UTR of eNOS mRNA [68]. Estrogen represents a different class of agents capable of modestly enhancing eNOS expression [69,70]. Clearly, in vivo, estrogen exerts both rapid non-genomic effects [71,72] and long term transcriptional actions on eNOS and more importantly, basal NO production.

2.11 Post-translational regulation

Optimal NO release from eNOS is known to depend on targeting of this enzyme to Golgi regions and to plasmalemal caveolae; events that require both N-myristoylation and cysteine palmitoylation of the enzyme [73–78]. The proper localization of eNOS is likely requisite for the ability of eNOS to interact with specific proteins that may regulate eNOS trafficking or activity. Recently, several proteins – other than calmodulin – have been demonstrated to interact with eNOS in co-precipitation experiments as well as biochemically. Negative regulation of eNOS occurs by binding to the coat protein of endothelial caveola, caveolin-1 [8,79] or to the intracellular domain (ID4) of G-protein coupled receptors (GPCR) [80]. Caveolin-1 negatively regulates NOS activity and NO production. This occurs via binding of the scaffolding domain of caveolin to a caveolin binding domain on eNOS [8,79,81]. Similar regulation occurs with GPCR. Conversely stimuli such as VEGF, histamine or fluid shear stress cause the recruitment of Hsp90 to eNOS resulting in activation of eNOS and NO release [82]. The dynamic interplay between caveolin, GPCR and Hsp90 regulatory proteins are not known.

Changes in the phosphorylation of eNOS have been observed following exposure to calcium mobilizing agonists [83], shear stress [84,85] and tyrosine phosphatase inhibitors [84,86]. In vitro, eNOS can be phosphorylated by a variety of kinases including protein kinase C and AMP kinase [87,88] while in vivo the kinases responsible have been elusive. Recently, we and others have shown that serine 1179 in bovine eNOS (or 1177 of human eNOS) is a site of phosphorylation in vitro and in vivo for the serine/threonine kinase Akt (also known as protein kinase B) [89,90]. Interestingly, the phosphorylation of the enzyme renders it more sensitive to calcium/calmodulin suggesting the introduction of the negative charge in the carboxy tail influences NOS catalytic function. This regulation appears to be critically important for both calcium-dependent and calcium-independent activation of eNOS by diverse stimuli.

3 General perspectives: the use of knockout strains to address the role of NO in the cardiovascular system

In the past 2 years or so, investigators have been exploring both old and new functional roles of NO using gene targeting to disrupt the NOS genes in mice. This genetic approach is very powerful permitting one to examine the basic roles of NO, without the use of pharmacologic agents, which can inhibit all NOS isoforms or exert other non-specific effects. Interestingly, all three NOS isoforms were successfully disrupted with no particularly obvious embryonic lethality occurring. This was somewhat disappointing and surprising given the fundamental roles for NO in a variety of basic signaling systems that are relevant during development and post-natal survival. Moreover, combination knockouts, to date, have not yielded any further information demonstrating an essential role for NO during the processes of embryogenesis.

With this in mind, how do we as investigators in the NO field rationalize the importance of NO in almost every system without an essential biological function? Many studies demonstrating a role for NO in biological systems do so by the use of NOS inhibitors bringing up the possibility that these excellent drugs are not as specific as once thought. Another relatively simple answer is that compensation for the loss of NOS and NO must occur very early in life so that the organism survives by controlled expression of other genes that are functionally redundant with the NO system. If this is the case, the presence of a redundant system ‘validates’ the importance of the NOS gene being deleted. Indeed there is evidence that other NOS isoforms may exert a biological role when one isoform is deleted [91–93]. An alternative explanation is that NO, under the most favorable conditions is an autacoid, therefore, a modulator of specific pathways in specialized cells. This theory is reasonable since many cells do not express a NOS isoform or produce NO in vivo or in vitro, but may express NOS once stressed or may respond to NO. In this scenario, one can imagine NO as an important, but non-essential regulator of signaling in specialized NO producing and responding cells. In fact, a majority of the phenotypes observed in NOS knockout mice occur post-natally, and support the idea that NO is critical for the ability of fully developed mice to acclimate to a given cellular response.

4 Phenotypes of isoform specific knockout mice

4.1 nNOS

After the identification of nNOS in central and peripheral nerves and the tremendous importance of this isoform in neurotransmission, learning, penile erection and ischemic injury based on animal and cell based studies, it was anticipated that nNOS knockout mice would possibly die in utero, be infertile and lack memory. As always the case in science, contrary to exceptions, nNOS knockouts were viable, fertile and able to learn the basics of survival. These mice lack nNOS protein expression in neuronal tissues and exhibit a pronounced, but not complete, loss in NOS catalytic activity in the nervous tissues examined [94]. This residual NOS activity is attributed to novel neuronal NOS isoforms generated by alternative splicing [20]. The only obvious pathological changes were grossly distended stomachs reminiscent of pyloric stenosis [94] and bladder–urethral sphincter dysfunction [95]. These effects were likely due to the importance of nNOS derived NO as a non-adrenergic, non-cholinergic transmitter in the peripheral nervous system [96,97].

4.2 Roles of nNOS in CNS

However, after more rigorous analysis, male nNOS knockout mice show increased aggressive behavior and excess inappropriate sexual behavior [98]. These behaviors were not associated with increased plasma levels of testosterone, increased anxiety or impaired olfactory function. The lack of effect of nNOS gene disruption on standard behaviors is not surprising since these mice exhibit normal long-term potentiation (LTP), a paradigm for memory and learning [91]. However, double knockout of nNOS and eNOS does reduce LTP in sub-populations of hippocampal neurons [92].

4.3 Stroke and ischemic injury

One to three days after middle cerebral artery (MCA) occlusion, infarct volumes and neurological deficits were reduced in nNOS knockout mice compared to littermate sibs [99]. In addition, administration of l-NAME to the nNOS knockout mice, to inhibit the vascular isoform eNOS, increased tissue damage. Similar results were obtained in isolated neurons from nNOS knockout mice [100]. These observations taken together with the data from experiments with eNOS knockout mice (see below), suggests neuronal NO exacerbates acute ischemic injury, whereas vascular NO preserves blood flow and protects against damage, thus providing a proof of concept that activation of signaling pathways in the brain generate a neurotoxic species of NO. The protective role of nNOS knockout during ischemic damage was also observed in response to neurotoxicity elicited by malonate and in isolated cortical neurons treated N-methyl-d-aspartate (NMDA) [101].

4.4 iNOS

A plethora of evidence is available for the expression of iNOS in many pathophysiological conditions both in humans as well as in animal models of disease. These include, but are not limited to, atherosclerosis, balloon injury and restenosis, cardiomyopathy, sepsis, transplant rejection and a variety of disorders associated with acute and chronic inflammation. Expression of iNOS is thought to have beneficial effects in some conditions due to its anti-microbial and anti-tumor activities, but may exert detrimental effects in other conditions such as septic shock. Mice with targeted disruption of the iNOS locus have been generated so far by three groups [102–104]. All of them have reported that the iNOS knockout mice are viable and indistinguishable from wild-type animals in gross appearance, reproduction, histology of all major organs and blood chemistries. As predicted, LPS is unable to induce iNOS in tissues from iNOS knockout mice.

4.5 Role of iNOS in inflammation, sepsis and vascular tone

Pharmacological experiments have implicated NO in local inflammatory responses, edema formation and sepsis [105,106]. Injection of carrageenin to the footpad of mice lacking iNOS was accompanied by a smaller degree of swelling than wild-type mice, suggesting that NOS mediates at least in part, immune inflammation [103]. Interestingly, after the administration of LPS to iNOS knockout mice, the adhesion and rolling of leukocytes in post capillary and post-sinusoidal capillaries is markedly enhanced suggesting that iNOS derived NO can exert cytoprotection as important negative regulator of leukocyte trafficking in the microcirculation [107].

In general, it is well recognized that NOS inhibition improves overall hemodynamics and lethality associated with LPS induced shock in rodents. Injection of LPS (1 mg/kg) into pentobarbital anesthetized wild-type mice reduced blood pressure rapidly and caused death within 5 h of injection. In contrast, heterozygote and homozygote knockout mice showed only a 30% and 15% decrease in blood pressure, respectively, and no lethality to the same dose of LPS suggesting that iNOS derived NO contributes to the actions of LPS [102]. However, plasma levels of tumor necrosis factor-α, interleukin-1β and interleukin-6, as well as indices of liver function were comparable in both genotypes while NO levels (measured as nitrite) were reduced in iNOS knockout mice. When conscious mice were injected with 5 or 15 mg/kg LPS, iNOS knockout mice were not protected against LPS lethality; however, a higher dose of LPS (30 mg/kg) was not as lethal in knockout mice, suggesting that iNOS deficiency under some conditions confers partial resistance to LPS [102]. An independent study did not confirm these observations, since iNOS knockout mice exhibited similar survival after LPS administration [104]. In yet a third study, iNOS null mice showed only a transient reduction in body mass and were completely resistant to the LPS (10 mg/kg)-induced lethality [103]. In a different model of septic shock, where conscious animals are injected with heat-inactivated P. acnes and LPS, iNOS knockout mice did not exhibit resistance to LPS death [102]. Although differences in the genetic background of the mice, the generation of knockout mice used and in the experimental conditions employed may contribute to these disparate findings, it is clear that iNOS derived NO is not the final common mediator of the hemodynamic collapse seen in sepsis.

During sepsis, the induction of iNOS in vascular smooth muscle is implicated as a causative factor in the hyperdynamic circulation accounting, in part, for the lowering of systemic blood pressure. As mentioned above, there is some evidence in support of this in vivo. A recent study in vitro has demonstrated that LPS reduces vasoconstrictor responses in isolated carotid arteries from wild-type mice; an effect completely absent in vessels from iNOS knockout mice [108]. Moreover, the ability of an iNOS inhibitor to restore vasomotor tone was not seen in vessels isolated from LPS injected iNOS knockout mice, consistent with a plethora of pharmacological studies documenting the role of iNOS in conduit vessel reactivity and tone.

4.6 Host defense mechanisms

iNOS derived NO is a double edge sword; it can participate in disease processes involving cytokine and immune activation and can act as an endogenous inhibitor in host defenses systems. This latter attribute is a testimony to the original work of Hibbs, Stuehr, Nathan and Marletta who discovered that NO is the principal cytotoxic substance produced from activated macrophages [109–112]. To directly test the role of iNOS is host defense, iNOS knockout mice were injected with Leishmania major and Listeria monocytogenes, respectively [102,103,113]. Wild-type and heterozygous mutant mice achieved spontaneous healing after injection with Leishmania promastigotes or Listeria whereas mutant mice were susceptible to infection and developed visceral disease. This data taken together, reinforce the important anti-bacterial and anti-tumor activities of iNOS-derived NO and suggests that isoform selective antagonism of iNOS may increase susceptibility to infections and dampen immunosurveillance mechanisms.

4.7 eNOS

The importance of endothelial derived NO as a regulator of cardiovascular homeostasis stemmed from the seminal observations that EDRF [114] is NO [115,116], NO relaxes blood vessels and inhibits platelet and leukocyte function and NOS inhibitors increase arterial pressure and reduce blood flow. Many of these biologic effects were confirmed in eNOS knockout mice.

4.8 Regulation of blood pressure and endothelium dependent vasorelaxation

eNOS knockouts have been generated by three groups independently and all report that homozygous knockout mice have elevated systemic blood pressures relative to control littermates [117–119]. eNOS knockout mice also have mild pulmonary hypertension [120]. An important ‘proof of concept’ developed in the eNOS knockout mice was the demonstration that acute injection of a l-arginine based NOS inhibitor no longer exerted a pressor response in vivo, but paradoxically reduced blood pressure [117]. The lack of a pressor response in eNOS knockouts genetically validates the pioneering studies that demonstrated the long lasting pressor effects of NG-mono methyl l-arginine (l-NMMA) on blood pressure [121,122]. The mechanism of the blood pressure elevation seen in knockout mice is still under question. The simple answer is that the absence of endothelial derived NO increases vascular tone causing increased vascular resistance and blood pressure, however, this has not been proven definitively. Previous studies have implicated NO in renin release which would modulate angiotensin II, a vasoconstrictor and critical regulator of blood pressure. Plasma renin levels were almost doubled in the eNOS mutants suggesting that angiotensin generation may contribute to the elevated blood pressure [118]. The reason for the depressor actions of a NOS inhibitor on blood pressure in eNOS knockout mice is not obvious but suggests that another NOS isoform (presumably nNOS) is responsible for the production of NO which modulates a pressor system (i.e. sympathetic nerves, arachidonic acid metabolic pathway) important for elevating systemic blood pressure. Recent evidence suggests that nNOS-derived NO is indeed the mediator. Using the selective nNOS inhibitor, 7-nitroindazole, blood pressure is reduced in eNOS mutant mice, whereas the inhibitor had a negligible effect in wild-type mice [123]. In addition, these authors show that the reduction in blood pressure in eNOS mutants was persistent and reversible when nNOS was inhibited chronically. This unexpected role of nNOS manifesting itself in an eNOS null background underscores the utility and importance of using genetically modified mice in physiological studies. A similar study in isolated pial arterioles showed that the nNOS inhibitor 7-nitroindazole had no effect on vascular responses in wild-type mice, but inhibited Ach induced relaxation in eNOS mutant mice, suggesting that eNOS mutants dilate pial arterioles via nNOS as a compensatory adaptive mechanism [10]. As predicted based on the hypertension seen in eNOS knockout mice, Ach induced vasorelaxation of mouse aortic rings was completely abrogated [117]. Similar results have been reported in the carotid arterial rings [124] demonstrating that eNOS derived NO is the endothelium derived relaxing factor as originally characterized by Furchgott.

4.8.1 Vascular remodeling and angiogenesis

Previous pharmacological studies have implicated eNOS in regulation of vascular smooth muscle remodeling in response to blood flow changes [125]. Recently this has been proven correct by Rudic et al. who showed that eNOS knockout mice failed to reduce lumen diameter in response to a reduction in blood flow [126]. This study demonstrated that NO was the endothelium-derived mediator necessary for chronic arterial remodeling [127]. One provocative finding in this study was the failure to reduce lumen diameter in eNOS knockout was associated with an increase in vessel wall thickness due to vascular smooth muscle hyperplasia suggesting that eNOS was a mechanosensor transducing hemodynamic changes into biochemical events in the underlying smooth muscle. In a similar study, Moroi et al showed that in response to a cuff injury placed around the femoral artery, eNOS mutant mice developed more severe intimal proliferation compared to wild-type mice [128]. Collectively, these two studies support the idea that endogenous NO is involved with the normalization of vessel shear stress and wall strain and is an important negative regulator of vascular smooth muscle cell proliferation.

Another facet of vascular remodeling in which NO has been implicated is angiogenesis. Indeed NO participates in the signaling pathway for the potent angiogenic factor, vascular endothelial growth factor (VEGF). NOS inhibitors block VEGF induced proliferation, migration and angiogenesis, in vitro and in vivo [129,130]. However, the role of endothelium derived NO is not essential for the process of vasculogenesis and angiogenesis during development as the mice survive and grow normally. This is in contrast to mice lacking VEGF- or VEGF-receptors which die early during vascular development [131]. Thus far, the only published data directly testing the role of NO in angiogenesis using eNOS knockouts is using a model of hind limb ischemia to trigger the angiogenic response. eNOS knockout mice exposed to hind limb ischemia did not re-vascularize the circulation as well as control mice [132]. This was documented by markedly lower blood flow into the ischemic region and a decrease in capillary density in knockout mice. Moreover, exogenous VEGF administration or VEGF gene therapy failed to restore angiogenesis in eNOS knockout mice, supporting the notion that NO is an essential downstream element regulating VEGF induced angiogenesis in adult mice. The implications of these findings are tremendous especially for physicians testing the clinical utility of VEGF for coronary angiogenesis.

5 Summary

Exciting new data elucidating the molecular biology of NOS regulation will no doubt lead to a greater understanding of how these genes are controlled in normal physiology and disease. The availability of NOS knockout mice will permit investigators to probe physiological and pathophysiological results of NOS deficiency disease. In this regard, the generation of tissue specific knockouts, as well as inducible knockout mice may reveal unappreciated insights into the functions of NO. Finally, recent evidence that nNOS and eNOS can interact with other proteins that may modulate the targeting or activity of these enzymes points toward the fascinating, yet subtle nature of how mammalian cells precisely control the production of NO.


This work was supported by grants from National Institute of Health (HL57665 and HL 61371). W.C.S. is an Established Investigator of the American Heart Association.


  • 1 Present address: Laboratory for Molecular Pharmacology, University of Athens, Athens, Greece.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
  51. [51]
  52. [52]
  53. [53]
  54. [54]
  55. [55]
  56. [56]
  57. [57]
  58. [58]
  59. [59]
  60. [60]
  61. [61]
  62. [62]
  63. [63]
  64. [64]
  65. [65]
  66. [66]
  67. [67]
  68. [68]
  69. [69]
  70. [70]
  71. [71]
  72. [72]
  73. [73]
  74. [74]
  75. [75]
  76. [76]
  77. [77]
  78. [78]
  79. [79]
  80. [80]
  81. [81]
  82. [82]
  83. [83]
  84. [84]
  85. [85]
  86. [86]
  87. [87]
  88. [88]
  89. [89]
  90. [90]
  91. [91]
  92. [92]
  93. [93]
  94. [94]
  95. [95]
  96. [96]
  97. [97]
  98. [98]
  99. [99]
  100. [100]
  101. [101]
  102. [102]
  103. [103]
  104. [104]
  105. [105]
  106. [106]
  107. [107]
  108. [108]
  109. [109]
  110. [110]
  111. [111]
  112. [112]
  113. [113]
  114. [114]
  115. [115]
  116. [116]
  117. [117]
  118. [118]
  119. [119]
  120. [120]
  121. [121]
  122. [122]
  123. [123]
  124. [124]
  125. [125]
  126. [126]
  127. [127]
  128. [128]
  129. [129]
  130. [130]
  131. [131]
  132. [132]
View Abstract