sign in sign up
<<
Home , Calcium in biology, Nitrate

Review Article Acta Cardiol Sin 2004;20:201-11

A Tale of Two Molecules: and Asymmetric Dimethylarginine

Ken Y Lin1,2 and Shih-Chung Lin3

Despite monumental progress in both molecular and interventional cardiology, atherosclerosis remains the leading cause of mortality and morbidity in the US, claiming more lives than the next 7 causes of mortality combined. Not only is this chronic disease the beginning of many familiar downstream clinical manifestations such as acute myocardial infarction, but research has demonstrated that the pathogenesis of atherosclerosis is intricately linked to many other cardiovascular abnormalities, from hyperhomocysteinemia to diabetes mellitus. Inflammation and oxidative stress, for instance, are common mediators for these diseases. Their interrelatedness has elicited much interest in finding a molecular marker with both prognostic and diagnostic values. Asymmetric dimethylarginine (ADMA) is one such molecule, recently gaining wider acceptance as a novel cardiovascular risk factor. As an endogenous competitive inhibitor of (NOS), ADMA is elevated in the plasma of patients with hypercholesterolemia, hypertension, hyperhomocysteinemia, , and insulin resistance. Not only is plasma ADMA concentration predictive of cardiovascular events, ADMA reduces the production of the anti-inflammatory and anti-atherogenic molecule, nitric oxide. As such, ADMA is both a disease marker and a mediator, whose pathophysiology merits attention. This article reviews the current knowledge of ADMA within the larger context of the NOS pathway.

Key Words: Nitric oxide · Asymetric dimethyl- · · Nitric oxide synthesis (NOS)

The endothelium nitric oxide (NO). The endothelium is a monolayer of that invests the lumen of all vessels. The normal endothelium Nitric oxide acts like a Teflon coat, preventing circulating blood ele- The pioneering work that led to the identification of ments from adhering to the vessel wall. Strategically NO as the endothelium-derived relaxing factor (EDRF) stationed between the smooth muscle cells and the circu- was no less celebrated than the molecule itself. As a lating blood, endothelial cells play a seminal role in the parasympathetic molecule, acetylcholine was known to intricate signal pathways that network our circulation to relax vascular smooth muscles, causing vasodilation. In- local tissues. To maintain these functions, endothelial triguingly, when a vessel ring was removed for ex vivo cells synthesize and respond to a plethora of molecules. studies in which acetylcholine was added, a contradic- Most famous among them in the past decade has been tory vasoconstriction was observed in the vessel section. Moreover, vascular constriction was observed if acetyl- choline was added directly to vascular smooth muscle Received: February 16, 2004 Accepted: June 4, 2004 cells. Robert Furchgott and colleagues explained these 1Division of Cardiovascular Medicine, Stanford University School of Medicine, Stanford, CA, USA. 2MD-PhD Program, Harvard Medical observations with a hypothetical EDRF, which was re- School, Boston, MA, USA. 3Lin Heart Clinic, Tainan, Taiwan. leased from endothelial cells upon their stimulation by Address correspondence and reprint requests to: Shih-Chung Lin, acetylcholine. Subsequent chemical characterization of MD, FACC, FCCM, President, Lin Heart Clinic, 403, Sec. 2, Min Sheng Road, Tainan, 703, Taiwan. Tel: 886-6-228-0808; EDRF by the laboratories of Louis Ignarro and Ferid E-mail: [email protected] Murrad led to the striking conclusion that NO, a com-

201 Acta Cardiol Sin 2004;20:201-11 KenYLinetal. mon air pollutant, is indeed the EDRF. For their atherosclerosis. As such, deficiency of NO increases contributions, Furchgott, Ignarro, and Murrad were vascular resistance and promotes atherogenesis;5 it also awarded the Nobel Prize in and Medicine in plays a key role in the progression of a constellation of 1998. cardiovascular abnormalities, including diabetes mellitus,6-8 Because its has an unpaired electron, NO is hypercholesterolemia,9 and hyperhomocysteinemia.10 a free radical. As such, its in vivo half-life is on the order of microseconds, the molecule being rapidly oxidized to Nitric oxide synthesis - (NO3 ). This short half-life gives NO just enough NO is converted from L-arginine by the NOS. time to diffuse from the endothelium, where the gas is The reaction involves the transfer of electrons from produced, to the neighboring smooth muscle cells, where NADPH, via the flavins FAD and FMN from the reductase the molecule induces its vasodilatory effect. Reduced NO region (C-terminal) of NOS to its oxidase region (N-termi- bioavailability occurs through 2 mechanisms: decreased nal), where L-arginine is converted to L-citrulline and production and/or increased degradation. In the former, NO.11 Other cofactors in this enzymatic reaction include nitric oxide synthase (NOS) makes less NO because of re- , zinc, and . In conditions duced substrate, cofactors, the enzyme itself, and/or where the substrate and/or cofactors become limited, the increased antagonists. In the latter, oxidative stress, char- electrons obtained from oxidizing NADPH would be di- - acterized by an increase in superoxide anion (O2 ), rected to , instead of L-arginine. Subsequently, quenches NO, turning it into (ONOO-), superoxide anion, rather than NO, is produced. Under this which is then oxidized to the inactive nitrate. condition, NOS is said to be “decoupled,” in the sense that Decreased NO bioavailability is the defining feature the enzyme’s reductase is short-circuited from the oxidase of endothelial dysfunction. In patients with atherosclero- region by molecular oxygen. In other words, depending on sis, endothelial dysfunction, indicated by reduced substrate availability, NOS would produce either NO or forearm blood flow following acetylcholine infusion, superoxide anion (Figure 1). can be observed before any morphological changes typi- Upon its release from NOS, NO diffuses across the en- cal of atherosclerosis. NO inhibits monocyte adhesion,1 dothelial and travels to the neighboring platelet aggregation,2,3 and vascular smooth muscle pro- smooth muscle cells, where the gas activates soluble liferation,4 all of which are key pathological processes in (sGC). The cytosolic sGC converts gua-

Figure 1. The nitric oxide synthase pathway. NOS generates either NO or superoxide anion, depending on the substrate availability. NO stimulates the enzyme soluble guanylate cyclase (sGC), which generates cGMP, leading to vascular relaxation. Superoxide anion quenches NO and inactivates it to nitrate. ADMA is an endogenous inhibitor of NOS and is itself degraded by the enzyme DDAH. Superoxide anion lowers the enzymatic activity of DDAH and thereby elevates the level of ADMA and reduces NO production.

Acta Cardiol Sin 2004;20:201-11 202 Nitric Oxide and Asymetric Dimethyl-Arginine nine triphosphate (GTP) to cyclic guanine monophosphate found to improve endothelial function by restoring NO (cGMP), which triggers a phosphorylation cascade that production.19,20 This phenomenon is generally known as leads to the relaxation of the smooth muscle cells. the arginine paradox. So far, no conclusive studies have There are three isoforms of NOS; namely, neuronal provided a satisfactory explanation. Recently, intracellular NOS (nNOS), inducible NOS (iNOS), and endothelial localization has been hypothesized to provide an insight NOS (eNOS). Both nNOS and eNOS are constitutively into the paradox. expressed, with the latter predominantly found in the cardiovascular system.12,13 It must be noted that the ex- NOS localization pression of eNOS is nonetheless subject to change. A The cell biology of eNOS is fascinating, and the variety of agents can augment eNOS expression. These field is still growing. eNOS is an N-myristoylated pro- include cyclosporin, glucose at high concentration, tein. N-myristoylation is a post-translational process low-density lipoprotein at low concentration, shear whereby a fatty acid chain is covalently attached to a stress, and estrogen. In fact, the promoter region of the protein, making it membrane-bound. This modification eNOS gene contains an estrogen response element, an is important for subcellular targeting of proteins to dis- observation that could in part explain estrogen’s crete microdomains of cells. Confocal imaging revealed cardioprotective effect. that eNOS is distributed on the cytoplasmic face of Increase in endothelial intracellular concen- Golgi and caveolae. Caveolae are specialized 50-100 nm tration following stimulation (such as acetylcholine) flask-shaped invaginations of the plasma membrane. activates eNOS because the enzyme requires the cal- Given that eNOS is localized, the aforementioned cium-bound calmodulin as a . Indeed, the intracellular arginine concentration, which is an average calcium-dependency of eNOS was confirmed by studies over the entire cell, might not necessarily reflect the lo- in which NO production and eNOS activity were both cal arginine concentration in the enzyme’s vicinity. abrogated after calcium removal.14,15 iNOS, on the other Emerging evidence, derived from our accumulating un- hand, is calcium-independent because it binds tightly to derstanding of caveolae, suggests that this is a valid calmodulin, and the binding interaction is irreversible; speculation. its regulation henceforth depends almost exclusively on The phospholipids that typically decorate the the transcriptional level, for which the enzyme receives plasma membrane are virtually absent in caveolae, the term “inducible.” which are enriched in certain lipid such as glycosphingolipids and structural proteins such as The arginine paradox caveolin. The fatty acid chains of these lipids tend to be NOS is an intracellular enzyme. Half-saturating extended and unsaturated, rendering the lipids more arginine concentration (Km) was reported as 1.4 mMfor tightly packed. Their tightness gives caveolae a higher nNOS, 2.83 mM for iNOS, and 2.9 mMforeNOS.16 ordered state in a typically fluid membrane. Accumulat- Freshly isolated endothelial cells have been found to con- ing evidence suggests that caveolae are sites where tain up to 2 mM of arginine.17 Given that the intracellular many pathways converge, because arginine concentration in endothelial cells is hundred-fold many signaling molecules, including eNOS, are found higher than the enzyme’s Km, eNOS should always be sat- in this specialized membrane region. Of note, colocalized urated. In other words, an increase in the substrate with eNOS in caveolae is the y+ transporter, a protein arginine’s concentration would not increase the produc- that transports cationic , including arginine, tion of NO. Intriguingly, numerous studies have indicated into the cell.21 Also clustered around eNOS are that intravenous or oral supplementation of arginine in argininosuccinate synthase and argininosuccinate lyase, humans augments endothelial NO production (a list of which are that convert L-citrulline, a by-prod- these studies can be found in the study by Flam et al).18 uct of NO production, back to L-arginine, the substrate Furthermore, in hypercholesterolemic humans and for NOS.22 In fact, a recent study showed that infusion animals, in whom NO-mediated endothelium-dependent of L-citrulline could also enhance NO production, mim- vasodilation is impaired, arginine supplementation was icking the effect of arginine infusion. Colocalization of

203 Acta Cardiol Sin 2004;20:201-11 KenYLinetal. these proteins with NOS suggests that the local arginine Accumulating evidence from both basic and clinical level around NOS could be highly regulated. Accord- investigations supports this concept. Vallance and col- ingly, the total intracellular arginine concentration leagues were among the first to document ADMA’s should be cautiously interpreted. clinical significance. They noted that in patients with Nevertheless, the hypothesis that eNOS localization end-stage renal disease, plasma ADMA accumulate due to answers the arginine paradox raises one critical question. reduced renal clearance.24 Reduced clearance of ADMA Membrane-bound eNOS is inactive. In caveolae, eNOS is in renal failure is associated with endothelial vasodilator associated with the protein caveolin-1, which prevents dysfunction, reversible by administration of L-arginine or calmodulin from binding to eNOS when intracellular cal- by dialysis, which removes plasma ADMA.25 cium concentration is low. The activation of eNOS requires, In subsequent years, several laboratories independ- in addition to a rise in calcium level, phosphorylation of ently demonstrated and confirmed the correlation between the enzyme by a serine kinase, and the enzyme’s elevated plasma ADMA and nearly all major cardiovas- depalmytoylation. This process releases eNOS from the cular risk factors such as aging,26 hypertension,27,28 membrane. That eNOS is inactive when localized and diabetes,29 insulin resistance,30 hypercholesterolemia, that the enzyme makes NO only when it leaves caveolae hypertriglyceridemia,31 and hyperhomocysteinemia.10 In raise a question regarding the extent to which eNOS lo- hypercholesterolemic humans, elevated plasma ADMA calization contributes to the arginine paradox. Further inversely correlated with endothelial-dependent vasodilation experiments are needed to answer this question. Here is in the forearm vasculature. Consistent with the idea of a an example using cell culture: Dr. William Sessa’s labo- competitive antagonism between ADMA and L-arginine, ratory at Yale University has developed eNOS mutants intravenous infusion of L-arginine restored endothelial that are kinetically indistinguishable from the wildtype.23 function and increased urinary nitrate excretion, which However, the mutants cannot be myristoylated; hence, served as a surrogate parameter of NO production. they remain cytosolic. After transfecting these mutants Plasma ADMA levels are also associated with other to endothelial cells, we could expose the cells to patho- cardiovascular complications such as ,32 conges- logical stimuli that are known to reduce NO production tive heart failure,33 and peripheral arterial disease (stimulus such as oxidized LDL). We could then apply (PAD).34 The level of ADMA was found to correlate arginine to the medium to determine whether increasing with the severity of atherosclerotic disease in patients the substrate would lead to greater amount of NO. If with PAD. Intravenous infusion of L-arginine both re- arginine fails to restore NO production in the transfected duced pain and improved walking distance. In addition, cells, we can conclude that eNOS localization does not ADMA levels have been shown to correlate with carotid contribute to the arginine paradox. intima-media thickness in patients with end-stage renal disease,35 and more strikingly, in a group of 116 appar- ADMA in cardiovascular diseases ently healthy individuals.26 While the precise role of eNOS localization in the Most recently, plasma concentrations of ADMA NO pathway still remains unsettled, another mechanism were reported a strong predictor of coronary events in for endothelial dysfunction in general, and arginine para- non-smoking male subjects with a history of coronary dox in specific, focuses on asymmetric dimethylarginine heart disease.36 Of note, patients in the upper quartile of (ADMA). As an endogenous, intracellular competitive ADMA plasma levels had a 3.9-fold enhanced risk of an antagonist of all isoforms of NOS,16 ADMA can reduce acute coronary event. NO synthesis if its concentration increases. In other Furthermore, Zoccali and colleagues recently pub- words, the level of ADMA could serve as a biochemical lished a large, prospective clinical study of 225 patients marker for NO availability. Given that NO deficiency is with end-stage renal disease undergoing hemodialysis the defining feature of endothelial dysfunction – the crit- (mean duration of hemodialysis 42 months).37 In this ical step in the onset of many cardiovascular diseases – study cohort, both fatal and non-fatal cardiovascular ADMA appears a promising candidate for cardiovascular events were recorded during a mean follow-up period of risk evaluation. 33.4 months. Strikingly, plasma ADMA levels emerged

Acta Cardiol Sin 2004;20:201-11 204 Nitric Oxide and Asymetric Dimethyl-Arginine as the second strongest predictor of all causes of mortal- ity after age, outweighing conventional risk factors such as hypertension, diabetes, hypercholesterolemia, and smoking. Commensurate with these human studies, animal and cell-culture works not only confirm the association between elevated ADMA and reduced NO, but also es- tablish the causal relation between the 2 molecules. Exogenously added ADMA in concentrations between 1 and 10 mmol/L dose-dependently reduces endothe- lium-dependent NO-mediated vasodilation in isolated rat mesenteric vessels38 and cerebral vessels,39 inhibits NO production by cultured macrophages, and increases en- dothelial adhesiveness to monocytes.1 Figure 2. Structure of arginine and endogenous methylarginines. ADMA is arguably the newest addition to an already Arginine is the substrate for nitric oxide synthase. Asymmetric sizable list of cardiovascular risk markers such as dimethylarginine (ADMA) and Monomethylarginine (NMMA) are fibrinogen, LDL, , interleukin-6, von Willebrand competitive inhibitors of NOS, while symmetric dimethylarginine factor, plasminogen activator inhibitor-1 and C-reactive (SDMA) is inactive toward the enzyme. protein. However, unlike some of these markers, ADMA is more than a marker; it plays a pivotal role in cardio- once. This configuration gives rise to symmetric vascular physiology by regulating the amount of NO dimethylarginine (SDMA).40 Both NMMA and ADMA synthesized. Hence, understanding the regulation of block the activity of NOS; SDMA, however, is inactive plasma ADMA elevation could lead to novel therapeutic to NOS, presumably because its guanidino group is too approaches and treatment modalities. Presently, the large to fit into the active site of NOS (Figure 2). mechanism of increased ADMA level is not fully charac- terized. Plasma ADMA level reflects a dynamic balance ADMA clearance between the molecule’s generation and its degradation. Renal excretion is one common way by which As such, increased ADMA is a consequence of increased ADMA, NMA, and SDMA are removed.41 This observa- generation and/or decreased clearance. tion might in part explain the rise in plasma ADMA in patients with renal failure. In that particular context, the ADMA generation kidney acts as a sink for ADMA. Once the sink has mal- In endothelial cells, ADMA is generated from the functioned, ADMA accumulates in the source - the endo- proteolysis of ubiquitous proteins containing methyl- thelial cells - where it reduces the production of nitric ated arginine residue. In the nucleus, transcriptional oxide. regulators are often methylated by enzymes called The degradation of ADMA and NMA, but not methyl transferases. In particular, protein arginine SDMA, occurs via an additional pathway. The enzyme methyl transferase (PRMT) transfers a to dimethylarginine dimethylaminohydrolase (DDAH) me- one of the guanidino on an arginine residue. tabolizes ADMA to dimethylamine and citrulline. Either one or two methyl groups can be added to Currently, two DDAH isoforms have been identified. arginine, subsequently resulting in monomethylarginine DDAH-I is found mostly in neural , while (NMMA) and dimethylarginine., respectively. In the DDAH-II predominates in peripheral and vascular tis- case of dimethylarginine, both methyl groups can oc- sues. DDAH is an intracellular protein. Its distribution cupy the same guanidino nitrogen, giving rise to does not seem to be localized within any intracellular asymmetric dimethylarginine (ADMA). Alternatively, compartments, nor is the enzyme membrane-bound. In- the 2 methyl groups can be distributed symmetrically terestingly, red blood cells also contain DDAH, such that each guanidino nitrogen is only methylated suggesting a possible role of erythrocytes in regulating

205 Acta Cardiol Sin 2004;20:201-11 KenYLinetal. plasma level of ADMA. ciation of this molecule’s clinical importance rests the The role of DDAH in a variety of disease states has question of how to quantify ADMA. There are two chal- been intensely investigated.8-10 Given the enzyme’s role lenges that any ADMA assay must overcome. First, the in removing ADMA, reduction of DDAH enzymatic ex- detection must be sufficiently sensitive to distinguish pression or activity would elevate ADMA and subsequently ADMA from its endogenous isomer, SDMA. Second, the reduce NO – the hallmark of endothelial dysfunction. assay must possess sub-micromolar sensitivity, because Indeed, John Cooke’s laboratory has shown that an impair- plasma level of ADMA is usually in the range of 0.5 to 3 ment of DDAH activity almost always accompanies a rise in mM. plasma ADMA. In fact, a 30-50% reduction in DDAH en- Virtually all of the laboratories around the world zyme activity was observed in hyperhomocysteinemia, currently investigating ADMA measure the molecule us- hypercholesterolemia, and hyperglycemia, both in vitro ing high pressure liquid chromatography (HPLC).8,43-45 and in vivo. In several papers, Cooke’s colleagues have Briefly, the method involves pre-concentration of also shown a strong negative correlation between DDAH ADMA using solid-phase extraction, with a strong cat- activity and plasma ADMA level.8 Taken together, this exchange column (available from Varian Inc., Palo line of evidence suggests that DDAH plays a key role in Alto, USA). Solid-phase extraction removes most of the regulating the concentration of ADMA. As such, it be- chemical components from the sample (which could be comes imperative to identify how DDAH activity is plasma, urine, or culture medium), leaving only moder- lowered. ately basic compounds, which include ADMA, SDMA, DDAH is an oxidant-sensitive enzyme. At its active and arginine.8 site, a cysteine residue initiates the first chemical reac- The concentrated sample is labeled with the tion that leads to formation of the enzymatic reaction fluorochrome, o-phthaldialdehyde or naphthalene-2, intermediate. This cysteine residue 249 is vulnerable to 3-dicarboxaldehyde.43 After the labeling (or the oxidative attack from many intracellular oxidants, in- “derivatization”), components within the sample are - 42 cluding homocysteine and superoxide anion (O2 ). An separated using reverse-phase HPLC with a nucleosil imbalanced cellular oxidative state might oxidatively phenyl column. A detector then quantifies modify the cysteine residue of DDAH. In the case of each sample component as it exits the column. homocysteine, a bond actually forms between HPLC separates compounds based on their structural the oxidant and the cysteine residue of DDAH.10 This dissimilarity, and the time a compound takes to exit the and other oxidative interaction with the cysteine residue column is in part determined by its relative hydrophobicity drastically lowers the activity of DDAH. Administration compared to other compounds present in the sample. Be- of antioxidants such as pyrollidine dithiocarbamate and cause ADMA is among the most hydrophobic chemicals conjugated superoxide dismutase in the concentrated sample, it is among the last com- have been shown to restore the enzymatic activity of pound to exit the column. The literature reports an DDAH in endothelial cell culture.8,10 elution time between 20 to 50 minutes for ADMA. The There is already a sizable list of pathways activated total time of the assay also includes the time spent on by oxidative stress that render cells in a pro-inflamma- solid phase extraction and on cleaning the column after tory state. The vulnerability of DDAH to oxidative stress each sample run. In sum, one sample takes at least 40 presents another mechanism whereby oxidative stress minutes to process. As such, HPLC is very time- and la- aggravates endothelial dysfunction. Pathological condi- bor-intensive, and is prone to human and mechanical tions that elevate oxidative stress lower the activity of errors. DDAH, leading to higher cellular and plasma levels of HPLC’s intrinsic variability notwithstanding, there ADMA, which in turn reduces the production of NO. is no unified and consistent protocol of measuring ADMA using HPLC. Different mobile phases, column ADMA assay in biological samples types, and derivatization times are usually employed in Literature on ADMA has been on a rapid rise over various laboratories. As such, there is often a significant the last few years. Concurrent with our increasing appre- difference between the control groups from 2 different

Acta Cardiol Sin 2004;20:201-11 206 Nitric Oxide and Asymetric Dimethyl-Arginine laboratories. come foam cells, setting the stage ready for the A study, initiated by Dr. Böger from Germany, is development of fatty streak and early lesions of athero- currently under its way to collect data from the major sclerosis. ADMA laboratories around the world and to compare ADMA is among most potent and abundant endoge- their differences. Results from this study would certainly nous inhibitors of eNOS. Today some doubts remain stimulate more communication and collaboration among regarding the relevance of ADMA to the NO pathway. laboratories investigating ADMA. Most of the critics of ADMA argue that ADMA’s plasma Other methods to quantify ADMA have also been level is too low, and the extent of its variation between published and proposed. A French group has reported normal and pathological condition is insignificant, to ef- successful separation and quantification of ADMA and fectively compete with arginine, which has a ten-fold SDMA using laser-induced fluorescence capillary elec- higher concentration in the plasma. The same ADMA- trophoresis (LIF-CE).46 This technique separates to-arginine ratio also holds true within the endothelial compounds based on difference in their charge-to-mass cells. However, as we have pointed out earlier, caution ratio. However, ADMA and SDMA have the same mo- must be taken when interpreting intracellular concentra- lecular mass, and their charge difference under any pH is tions, because the numbers reflect only the average, but so subtle that separation would be difficult under any not the local concentration of ADMA and arginine, buffer system. In fact, 2 groups from Stanford University which, given the localization of eNOS, are biochemically have failed to measure ADMA using the reported more relevant. LIF-CE protocol, but are currently working on a modi- Cellular compartmentalization theory addresses the crit- fied protocol that introduces organic additive into the ics’ concerns; however, it does so by simply citing the buffer system to enhance the electrochemical difference unknown – the cell biology of eNOS, ADMA, and arginine. among the chemicals (unpublished data). Several private Another more direct piece of evidence supporting the role of firms have also started developing immunoassay for de- ADMA comes from a standard practice in a pharmacology tecting ADMA. Such an approach relies on an laboratory. Monomethylarginine (L-NMMA) and L-ni- ADMA-specific antibody to identify and quantify the tro-arginine methylester (L-NAME) are 2 commonly used molecule. If realized, this technique would enable agents to induce systemic vasoconstriction in humans and ADMA detection using enzyme-linked immuno-sorbent laboratory animals. They are usually administered with a assay (ELISA), which offers both high sensitivity (at the low dose that would elevate their plasma concentrations 3- level of nanomole) and high throughput (96 samples in a to 4-fold, which are still below 5 mM.Thesystemiceffects few seconds). of these eNOS inhibitors have been well documented for The field of ADMA is a wonderful example of how many years. Given that they share similar pharmacokinetic interdisciplinary approach could accelerate and catalyze profile with ADMA, ADMA should likewise affect eNOS at progress in today’s medicine. The number of analytical micromolar quantity. More importantly, this line of logic chemistry laboratories looking at ways to improve suggests that eNOS’ access to arginine might not be as ADMA measurement is growing as rapidly as the num- straightforward as vascular biologists once thought. Many ber of clinical and basic laboratories investigating the factors could upset arginine’s availability to eNOS, from a pathophysiology of ADMA. These trends only under- decrease in arginine itself (possibly caused by increased score ADMA’s diagnostic and prognostic values. arginine turnover by ) to an increase in eNOS inhib- itor such as ADMA, which could be caused by oxidative ADMA: a critical appraisal stress and DDAH dysfunction. NO plays a critical role in maintaining vascular tone If the level of ADMA is clinically important, the fac- and integrity. When NO becomes deficient, the endothe- tors that influence DDAH enzymatic activity are also lium becomes adhesive, attracting monocytes and other critical, for the latter directly regulates the former. Ac- blood elements to adhere and infiltrate through the endo- cordingly, a clinical study has been launched in Europe thelium. In the presence of oxidized LDL in the to investigate the relation between DDAH single nucleo- sub-enthdothelial subspace, the infiltrated monocytes be- tide polymorphisms and a variety of cardiovascular

207 Acta Cardiol Sin 2004;20:201-11 KenYLinetal. diseases. In the preliminary data, DDAH actually MD-PhD student at Harvard Medical School; he receives emerged a stronger predictor for the onset of Type 2 dia- a fellowship from the Medical Scientist Training Program betes than markers such as triglyceride and . of the National Institution of Health. Meanwhile, the laboratory of John Cooke at Stanford University has also developed transgenic mice over-ex- pressing human DDAH I gene. Recently, we have shown REFERENCES that acute peritoneal injection of ADMA instantly at- 1. Boger RH, Bode-Boger SM, Tsao PS, et al. An endogenous tenuates insulin sensitivity in mice.47 Meanwhile, the inhibitor of nitric oxide synthase regulates endothelial adhesiveness insulin sensitivity of transgenic mice over-expressing for monocytes. J Am Coll Cardiol 2000;36:2287-95. DDAH is not significantly different from that of mice re- 2. Tsao PS, Theilmeier G, Singer AH, et al. L-Arginine attenuates ceiving the control treatment. Presumably, this is due to platelet reactivity in hypercholesterolemic rabbits. Arterioscler the transgenic mice’s enhanced ability to degrade the ex- Thromb 1994;14:1529-33. ogenously added ADMA. Taken together, this line of 3. Shukla SD, Paul A, Klachko DM. Hypersensitivity of diabetic initial evidence suggests that DDAH dysfunction in par- human platelets to platelet activating factor. Thromb Res 1992; 66:239-46. ticular, and very possibly endothelial dysfunction in 4. Ignarro LJ, Buga GM, Wei LH, et al. Role of the arginine-nitric general, could lead to insulin resistance and Type 2 dia- oxide pathway in the regulation of vascular smooth muscle betes. A recent clinical study corroborated this hypothesis proliferation. Proc Natl Acad Sci 2001;98(7):4202-8. by showing that eNOS polymorphism is associated with 5. Cooke JP and Dzau VJ. Nitric oxide synthase: role in the genesis Type 2 diabetes and insulin resistance syndrome.48 of vascular disease. Ann Rev Med 1997;48:489-509. While this concept still requires more rigorous proof, 6. Van de Ree MA, Huisman MV, de Man FH, et al. Impaired these studies embody the now decades-old belief that endothelium-dependent and independent vasodilation in type 2 diabetes mellitus and the lack of effect of . Cardiovas metabolic abnormality is also a vascular disease. Res 2001;52:299-305. The etiology of atherosclerosis is certainly multifac- 7. Tesfamariam B, Cohen RA. Free radicals mediate endothelial cell eted. ADMA and DDAH might also have yet-unknown dysfunction caused by elevated glucose. Am J Physiol 1992;263: interactions with other components of the NOS and in- H321-6. flammatory pathways, such as interleukin-8, protein 8. Lin KY, Ito A, Asagami T, et al. Impaired nitric oxide synthase kinase C and phosphoinositol-3-kinase. Treatments that pathway in diabetes mellitus: role of asymmetric dimethylarginine aim to restore 1 pathway might bestow the benefits at the and dimethylarginine dimethylaminohydrolase. Circulation 2002; 106:987-92. expense of another pathway. Hence, future studies 9. Boger RH, Bode-Boger SM, Szuba A, et al. Asymmetric should be directed to untangling these intricate interac- dimethylarginine (ADMA): a novel risk factor for endothelial tions and based on these mechanistic understanding, to dysfunction: its role in hypercholesterolemia. Circulation 1998; design better clinical guidelines and therapeutic inter- 98:1842-7. ventions. 10. St linger MC, Tsao PS, Her J-H, et al. Homocysteine impairs the NO synthase pathway: role of ADMA. Circulation 2001;104: 2569-75. 11. Iyengar R, Stuehr DJ, Marletta MA. Macrophage synthesis of ACKNOWLEDGEMENTS , nitrate, and N-: precursors and role of the respiratory burst. Proc. Natl. Acad. Sci 1987;84:6369-73. The authors of this paper would like to thank Dr. 12. Vanhoutte PM. Endothelial dysfunction and atherosclerosis. Eur John P. Cooke, who directs Stanford University Medical Heart J 1997;87:E19-29. School’s Division in Cardiovascular Medicine and who 13. Harrison DJ. Cellular and molecular mechanisms of endothelial has mentored and inspired Ken Lin during his undergrad- cell dysfunction. J Clin Invest 1997;100:2153-7. uate years. Ken Lin is an American Heart Association 14. Busse R and Mulsch A. Calcium-dependent nitric oxide synthesis in endothelial is mediated by calmodulin. FEBS Lett Western Affiliates Undergraduate Research Fellow, and a 1990;265:133-6. recipient of the Howard Hughes Medical Institute’s Un- 15. Forstermann U, Pollock JS, Schmidt HH, et al. Calmodulin- dergraduate Grant. These 2 institutions have in part dependent endothelium-derived relaxing factor/nitric oxide supported this study. Ken Lin is currently a first-year synthase activity is present in the particulate and cytosolic

Acta Cardiol Sin 2004;20:201-11 208 Nitric Oxide and Asymetric Dimethyl-Arginine

fractions of bovine aortic endothelial cells. Proc Natl Acad Sci 31. Lundman P, Erikson MJ, St linger MC, et al. Mid-to-moderate 1991;88:1788-92. hypertriglyceridemia in young men is associated with endothelial 16. Forstermann U, Closs EI, Pollock JS, et al. Nitric oxide synthase dysfunction and increased plasma concentrations of asymmetric isozymes: characterization, purification, molecular cloning, and dimethylarginine. J Am Coll Cardiol 2001;38:111-6. functions. Hypertension 1994;23:1121-31. 32. Yoo JH and Lee SC. Elevated levels of plasma homocysteine and 17. Hecker M, Mitchell JA, Harris HJ, et al. Endothelial cells asymmetric dimethylarginine in elderly patients with stroke. metabolize NG-monomethyl-L-arginine to L-citrulline and Atherosclerosis 2001;158:425-30. subsequently to L-arginine. Biochem Biophys Res Commun 1990; 33. Usui M, Matsuoka H, Miyazaki H, et al. Increased endogenous 167:1037-43. nitric oxide synthase inhibitor in patients with congestive heart 18. Flam BR, Hartmann PJ, Harrell-Booth M, et al. Caveolar failure. Life Sci 1998;62:2425-30. localization of arginine regeneration enzymes, argininosuccinate 34. Boeger RH, Bode-Boeger SM, Thiele W, et al. Restoring vascular synthase, and lyase, with endothelial nitric oxide synthase. Nitric nitric oxide formation by L-arginine improves the symptoms of Oxide 2001;5:187-97. intermittent claudication in patients with peripheral arterial 19. Girerd XJ, Hirsch AT, Cooke JP, et al. L-arginine augments occlusive disease. J Am Coll Cardiolol 1998;32:1336-44. endothelium-dependent vasodilation in cholesterol-fed rabbits. 35. Zoccali C, Benedetto FA, Maas R, et al. Asymmetric Circ Res 1990;67:1301-8. dimethylarginine, C-reactive protein, and carotid intima-media 20. Drexler H, Zeiher AM, Meinzer K, et al. Correction of endothelial thickness in end-stage renal disease. J Am Soc Nephrol 2002; dysfunction in coronary microcirculation of hypercholesterolemic 13:490-6. patients by L-arginine. Lancet 1991;338:1546-50. 36. Valkonen VP, Paiva H, Salonen JT, et al. Risk of acute coronary 21. Closs EI, Basha FZ, Habermeier A, et al. Interference of events and concentration of asymmetrical dimethylarginine. L-arginine analogues with L-arginine transport mediated by the Lancet 2001;358:2127-8. y+ carrier hCAT-2B. Nitric Oxide 1997;1:6573. 37. Zoccali C, Bode-B er S, Mallamaci F, et al. Plasma Concentration 22. Vallance P, Leone A, Calver A, et al. Accumulation of an of asymmetrical dimethylarginine and mortality in patients with endogenous inhibitor of nitric oxide synthesis in chronic renal end-stage renal disease: a prospective study. Lancet 2001;358: failure. Lancet 1992;339:572-5. 2113-7. 23. Fulton D, Fontana J, Sowa G, et al. Localization of endothelial 38. Kurose I, Wolf R, Grisham MB, et al. Effects of an endogenous nitric oxide synthesis phosphorylated on serine 1179 and nitric inhibitor of nitric oxide synthesis on postcapillary venules. Am J oxide in Golgi and plasma membrane define the existence of 2 Physiol 1995;268:H2224-31. pools of enzymes. J Biol Chem 2002.8;277(6):4277-84. 39. Faraci FM, Brian JE, Heistad DD. Response of cerebral blood 24. Vallance P, Leone A, Calver A, et al. Accumulation of an vessels to an endogenous inhibitor of nitric oxide synthesis. Am J endogenous inhibitor of nitric oxide synthesis in chronic renal Physiol 1995;269:H1522-7. failure. Lancet 1992;339:572-5. 40. B er RH, Sydow K, Borlak J, et al. LDL cholesterol upregulates 25. Kielstein J, Boeger R, Bode-Boeger S, et al. Asymmetric synthesis of asymmetrical dimethylarginine in human dimethylarginine plasma concentrations differ in patients with endothelial cells: involvement of S-adenosylmethionine-dependent end-stage renal disease: relationship to treatment method and methyltransferases. Circ Res 2000;87(2):99-105. atherosclerotic disease. J Am Soc Nephrol 1999;10:594-600. 41. Jacobi J, Schneider MP, John S, et al. Impact of NO-synthase 26. Miyazaki H, Matsuoka H, Cooke JP, et al. Endogenous nitric inhibition on renal hemodynamics in normotensive and hypertensive oxide synthase inhibitor: a novel marker of atherosclerosis. subjects. J Hypertens 2002;3:525-30. Circulation 1999;99:1141-6. 42. Murray-Rust J, Leiper J, McAlister M, et al. Structural insights 27. Surdacki A, Nowicki M, Sandmann, et al. Reduced urinary into the of cellular nitric oxide synthase inhibitors by excretion of nitric oxide metabolites and increased plasma levels dimethylarginine dimethylaminohydrolase. Nat Struct Biolo of asymmetric dimethylarginine in men with essential hypertension. 2001;8:679-83. J Cardiovasc Pharmacol 1999;33:652-8. 43. Marra M, Bonfigli AR, Testa R, et al. High-performance liquid 28. Fujiwara N, Osanai T, Kamada T, et al. Study on the relationship chromatographic assay of asymmetric dimethylarginine, symmetric between plasma nitrite and nitrate level and sensitivity in dimethylarginine, and arginine in human plasma by derivatization human hypertension: modulation of nitric oxide synthesis by salt with naphthalene-2,3-dicarboxaldehyde. Anal. Biochem 2003; intake. Circulation 2000;101:856-61. 318(1):13-7. 29. Abbasi F, Asagami T, Cooke JP et al. Plasma concentrations of 44. Teerlink T, Nijveldt RJ, de Jong S, et al. Determination of arginine, asymmetric dimethylarginine are increased in patients with type 2 asymmetric dimethylarginine, and symmetric dimethylarginine in diabetes mellitus. Am J Cardiol 2001;88:1201-3. human plasma and other biological samples by high-performance 30. St linger MC, Abbasi F, Chu JW, et al. Relationship between liquid chromatography. Anal. Biochem 2002;303(2):131-7. insulin resistance and an endogenous nitric oxide synthase 45. Pi J, Kumagai Y, Sun G, et al. Improved method for simultaneous inhibitor. JAMA 2002;287:1420-6. determination of L-arginine and its mono- and dimethylated

209 Acta Cardiol Sin 2004;20:201-11 KenYLinetal.

metabolites in biological samples by high-performance liquid 47. Lin KY and Cooke JP. Acute administration of asymmetric chromatography. J Chromatogr B Biomed Sci Appl 2000;742(1): dimethylarginine reduces insulin sensitivity in mice [abstract 199-203. submitted to American Heart Association 2003 Scientific 46. Causse E, Siri N, Arnal JF, et al. Determination of asymmetrical Sessions]. dimethylarginine by capillary electrophoresis-laser-induced 48. Monti LD, Barlassina C, Citterio L, et al. Endothelial nitric oxide fluorescence. J Chromatogr B Biomed Sci Appl 2000;741(1): synthase polymorphisms are associated with type 2 diabetes and 77-83. the insulin resistance syndrome. Diabetes 2003;52(5):1270-5.

Acta Cardiol Sin 2004;20:201-11 210 Review Article Acta Cardiol Sin 2004;20:201−11

雙分子傳奇：一氧化氮與非對稱性二甲基精氨酸

林耕仰 1,2 林世崇3 美國史丹佛大學 醫學院 心臟血管科1 美國麻州哈佛大學 醫學院 MD. Ph. D 博士班2 台灣台南市 林世崇心臟內科診所3

雖然在分子心臟學與介入心臟學巨大的進步，動脈粥狀硬化仍在美國居患病率及死亡率的 首要原因，且遠超過排行其下七項疾病之總數。動脈粥狀硬化為慢性疾病，它不僅是許多 眾所周知的臨床疾病的初始，例如急性心肌梗塞，研究亦顯示血管硬化的病理也與其他心 血管疾病，從高同氨酸症到糖尿病，都有錯綜複雜的關係。例如發炎反應，或氧化傷害， 即為這些疾病的共通媒介。其中的一氧化氮 (NO) 與非對稱性二甲基精氨酸 (ADMA) 在 血管硬化的病理變化過程中，扮演了極重要之角色。當發生發炎反應，或氧化傷害時，中 間的媒介分子不只使血管產生病變，同時也扮演了預後的因子。由於 ADMA 具有診斷及 預後的指標，因此最近被廣泛接受成為心臟血管疾病之新發現的危險因子之一。 ADMA 是 體內一氧化氮合成酵素 (NOS) 的內源性抑制分子，因此在高血脂、高血壓、高同氨酸症、 高血糖及胰島素阻抗的情況之下，血清中的 ADMA 值是偏高的。 ADMA 會減少抗發炎及 抗硬化分子的產生，因此也同時具有診斷及預後之價值。本文將探討 ADMA、NOS 生化 路徑及相互關係之最新研究。

關鍵詞：一氧化氮、非對稱性二甲基精氨酸、一氧化氮合成酵素、內皮細胞。

211