|Home | About | Journals | Submit | Contact Us | Français|
Soluble guanylate cyclase (sGC) is a key signal-transduction enzyme activated by nitric oxide (NO). Impaired bioavailability and/or responsiveness to endogenous NO has been implicated in the pathogenesis of cardiovascular and other diseases. Current therapies that involve the use of organic nitrates and other NO donors have limitations, including non-specific interactions of NO with various biomolecules, lack of response and the development of tolerance following prolonged administration. Compounds that activate sGC in an NO-independent manner might therefore provide considerable therapeutic advantages. Here we review the discovery, biochemistry, pharmacology and clinical potential of haem-dependent sGC stimulators (including YC-1, BAY 41-2272, BAY 41-8543, CFM-1571 and A-350619) and haem-independent sGC activators (including BAY 58-2667 and HMR-1766).
Nitric oxide (NO) is a key signalling molecule that is involved in the regulation of a variety of biological and physiological processes in mammals. Vast experimental and clinical evidence indicates that reduced bioavailability and/or responsiveness to endogenously produced NO contributes to the development of cardiovascular, pulmonary, endothelial, renal and hepatic diseases, as well as erectile dysfunction. Some of these conditions are currently treated with organic nitrates (such as glyceryl trinitrate) and other NO-donor or ‘nitrovasodilator’ drugs that release NO by spontaneous decomposition1 or bioconversion2 to activate soluble guanylate cyclase (sGC). However, the use of such compounds is limited by a potential lack of response due to insufficient biometabolism3, development of tolerance following prolonged administration4 and nonspecific interactions of NO with other biological molecules, including peroxynitrite-mediated tyrosine nitration5. The latter reactions are difficult to control owing to the spontaneous release of NO from nitrovasodilators and its free diffusion in biological systems. Moreover, despite the symptomatic improvement in patients with cardiovascular disease treated with organic nitrates, there is no clear evidence that such treatment reduces mortality. Therefore compounds that activate sGC in an NO-independent manner might offer considerable advantages over current therapies (FIG. 1). In this article, we describe the NO–sGC–cGMP signalling pathway and explain the therapeutic rationale for sGC stimulators and activators. The discovery, biochemistry and pharmacology of several such compounds are reviewed, and their potential for the treatment of a range of diseases, including arterial and pulmonary hypertension, heart failure, atherosclerosis, thrombosis, erectile dysfunction, renal fibrosis and failure, and liver cirrhosis, is discussed.
Guanine nucleotidyl (guanylyl; guanylate) cyclases (GCs) are widely distributed signal-transduction enzymes that, in response to various cellular stimuli, convert GTP into the second messenger cyclic GMP. The biological effects of cGMP are mediated by three major types of intracellular effectors: cGMP-dependent protein kinases I and II, cGMP-gated ion channels and cGMP-regulated phosphodiesterases (PDEs)6,7. Degradation of cGMP is catalysed by several differentially expressed PDE families (PDE1, 2, 3, 5, 6, 9, 10 and 11), which represent independent drug discovery targets7.
In contrast to the transmembrane particulate GC (pGC), which serves as a receptor for atrial, B-type and C-type natriuretic peptides, sGCs are receptors for gaseous ligands, namely NO and carbon monoxide (CO). They can associate with the plasma membrane through protein–protein interactions in a Ca2+-dependent manner8,9 or apparently in a constitutive manner10,11. Soluble GC is typically found as a heterodimer, consisting of a larger α-subunit and a smaller haem-binding β-subunit, although homodimers of these subunits can also form12. Four human sGC subunits exist: α1, α2, β1 and β2 of which the α1/β2 and α2/β1 heterodimers are the best characterized13,14. The 619-residue β-subunit contains an evolutionarily conserved amino-terminal haem-binding domain that has a length of about 200 residues15–17. A prosthetic haem moiety, which is crucial for the sensing of NO, is positioned in the haem-binding domain via its interaction with the axial ligand histidine-105 and the anchoring residues of the haem propionates tyrosine-135, serine-137 and arginine-139 (constituting the haem-binding motif Y-x-S-x-R; see FIG. 2)17–19. Based on sequence homology with the crystallized catalytic domains of adenylate cyclase, the carboxy-terminal catalytic domains of both sGC subunits are assumed to be orientated in a head-to-tail fashion20,21. The catalytic domains of both subunits are required for the formation of a catalytic active centre13,22,23.
The NO–sGC–cGMP pathway is crucial for the control of many physiological processes, such as host defence reactions, cell growth and proliferation, vascular homeostasis and neuronal transmission (FIG. 1)24. Although sGC is activated by nanomolar concentrations of NO, the sGC–cGMP axis might not be the only signalling pathway affected by NO. As a result of its reactivity with iron-containing catalytic sites, including the haem groups of FeS clusters, the function of various enzymes can be affected by NO, especially at higher concentrations. These alternative mechanisms could explain the distinct functional consequences of NO and cGMP signalling that have been reported, such as pro-25 and anti-aggregatory26,27 effects in platelets, and pro-28,29 and anti-atherosclerotic30,31 effects in blood vessels.
One major prerequisite for the NO-induced activation of sGC is the presence of the reduced Fe2+ haem moiety; its removal abolishes any NO-induced enzyme activation32,33. The central ferrous iron of the prosthetic group is coordinated between the four haem nitrogens and the axial ligand histidine-105, building a penta-coordinated histidyl-haem complex (FIG. 2). Binding of NO to this complex results in the formation of a hexa-coordinated histidine–haem–NO intermediate that rapidly decays into a penta-coordinated nitrosyl–haem complex. The cleavage of the haem–histidine bond is the molecular switch that leads to a ~200-fold activation of sGC34. However, this simple model of sGC activation does not fit with some subsequently observed activation characteristics of the enzyme. For example, several research groups have demonstrated that the conversion of the hexa-coordinated intermediate state into the penta-coordinated active nitrosyl–haem complex depends on the concentration of free NO, suggesting that there is a second binding site for NO35,36. Others have confirmed this finding, and reported that spectroscopically validated, NO-bound sGC can exist in a virtually inactive form37,38. This state, which is formed in the presence of low NO concentrations, can be transformed to the fully activated enzyme by addition of NO, the sGC substrate GTP, or the reaction products cGMP and/or pyrophosphate37,38. Physiological concentrations of ATP inhibit this transformation, linking NO–sGC–cGMP signalling to cellular energy metabolism37.
In addition to removal of the haem, its oxidation39 by sGC inhibitors such as ODQ (1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one), NS2028, methylene blue or ferri-cyanide leads to the formation of an NO-insensitive form of the enzyme40–45. Endogenously, changes in the redox state of sGC can be induced by reactive oxygen and nitrogen species such as superoxide (•O2−) and peroxynitrite (ONOO−), which are generated under conditions of oxidative stress46 (FIG. 1). Indeed, the capability of the prosthetic group of sGC to exist in different redox states might allow sGC, in addition to its gas-sensing capability, to monitor intracellular redox homeostasis (FIG. 3). With respect to disease, NO–sGC–cGMP signalling can be compromised either by reducing the bioavailability of NO (for example, via the chemical interaction of NO with •O2−) or by altering the redox state of sGC itself, making it unresponsive to endogenous NO and therefore to exo genous NO from NO-releasing drugs46. Excitingly, two novel drug classes seem to be able to overcome these obstacles: sGC stimulators and sGC activators (TABLES 1,,2).2). The former compounds stimulate sGC directly and enhance the sensitivity of the reduced enzyme to low levels of bio-available NO. Conversely, sGC activators do not modulate NO signalling at all but activate the NO-unresponsive, haem-oxidized or haem-free enzyme (FIG. 3). Clinically, the main rationale for these new therapeutic principles is based on both insufficient NO–sGC–cGMP signalling often associated with the use of conventional NO donors and the medical need to treat conditions associated with oxidative stress.
Our knowledge of the function of sGC has increased tremendously during the past decade. The recent discoveries of compounds that stimulate or activate sGC independently of NO release allow this venerable pharmacological target to be approached from a completely different perspective. NO-independent but haem-dependent stimulators of sGC, as well as NO- and haem-independent sGC activators, are emerging as valuable tools that could help to elucidate the physiology and pathophysiology of the NO–sGC–cGMP pathway in more detail. Given the clinical importance of diseases that these compounds could be used to treat, and on the basis of in vitro, in vivo and initial clinical studies, expectations for these novel drug classes are high.
In 1994, scientists at Bayer HealthCare AG started a search for substances that could induce an increase in NO synthesis and thereby stimulate sGC in porcine endothelial cells, in studies that involved the measurement of cGMP levels by radioimmunoassay. Nearly 20,000 compounds were tested, which led to the unexpected discovery of 5-substituted-2-furaldehyde-hydrazone derivatives as direct NO-independent sGC stimulators. Interestingly, the potency of these compounds in tests involving the stimulation of purified sGC and the relaxation of isolated blood vessels increased when they were exposed to light, which was clearly an unwanted feature for further drug development. In parallel, in December 1994, Ko and co-workers described a structurally related compound, the indazole derivative YC-1, which was subsequently characterized as an NO-independent, haem-dependent stimulator of highly purified sGC. Furthermore, its potency was unaffected by the prevailing light conditions47–50. During the following decade, various compounds that activate sGC in an NO-independent fashion were identified47,51–55. These drugs, including BAY 41–2272, BAY 41–8543, CMF-1571 and A-350619, constitute a novel class of haem-dependent sGC-stimulators, based on several shared characteristics: crucial dependency on the presence of the reduced prosthetic haem moiety and strong synergistic enzyme activation when combined with NO (TABLE 1).
YC-1 activates sGC independently of NO48–50,56. However, similarly to NO, the activating effect of YC-1 crucially depends on the presence of the reduced prosthetic haem moiety of sGC. Its removal or oxidation virtually abolishes any YC-1-induced sGC activation19,32,34,42–45,57–59. In the presence of YC-1, sGC activation by NO is potentiated. This effect is based, at least partially, on the strong stabilization of the nitrosyl-haem complex60 and the postulated transformation of the NO-activated enzyme from a low- to a high-output activation state37,38. Furthermore, YC-1 was shown to potentiate the weak sGC activator CO to a level of effectiveness comparable to NO48,49,61,62, and to inhibit cGMP-metabolizing PDEs63.
The exact mechanism of direct YC-1-induced effects on sGC and its putative binding site are still a matter of debate. Guided by homology between the catalytic domains of adenylate cyclases and GCs20 and the results of mutational studies64,65, it was suggested that YC-1 might bind to the catalytic domain of sGC, activating the enzyme by a mechanism that is comparable to forskolin-induced activation of adenylate cyclases13,21. This hypothesis recently gained support from the demonstration that YC-1 and related compounds interact with the catalytic subunit of sGC66. By contrast, further spectroscopic studies demonstrated that YC-1-induced effects, such as alterations of the CO–haem bond, could still be observed in haem-containing fragments of the β-subunit in the absence of the catalytic domain62. As there is a vital crosstalk between the mechanisms of NO- and YC-1-mediated sGC activation, this hypothesis is attractive. However, at the present time, both postulated binding mechanisms could be possible and will have to be validated in the future by co-crystallization studies.
BAY 41–2272 and BAY 41–8543 were synthesized based on YC-1 as a lead structure54,57,67,68. However, BAY 41–2272 and BAY 41–8543 are about two orders of magnitude more potent than YC-157. Both compounds activate purified sGC and strongly synergize with NO, reflecting the stabilization of the nitrosyl–haem complex57,67,69,70. A recent in vitro study using concentrations several orders of magnitude above those needed for sGC stimulation suggested some degree of PDE5 inhibition by BAY 41–227271. However, this effect is probably irrelevant because these high drug concentrations will not be reached therapeutically69,72,73.
Removal of the prosthetic haem moiety or its oxidation by ODQ strongly diminished BAY 41–2272/ 41–8543-induced enzyme activation. Using a photoaffinity-labelling approach based on a high-affinity analogue of BAY 41–2272, Stasch and co-workers observed labelling of cysteine-238 and -243 of the sGC α-subunit57. This result was validated with a photolabile derivative of YC-174. However, as no crystal structure of sGC is available, it is possible that these residues might be situated in the interface between the sGC subunits, thereby allowing labelling of the α-subunit even if the compound binds to the corresponding β-subunit, as discussed for YC-1. This view is supported by subsequent studies involving mutations in the α-subunit, which showed no effect on the BAY 41–2272-induced enzyme activation (P.M.S., unpublished data).
BAY 41–2272 and BAY 41–8543 produce potent relaxation of isolated systemic and coronary arteries and veins, and reduce coronary perfusion pressure in the rat heart Langendorff preparation. Both compounds also have antiproliferative properties in smooth muscles and anti-aggregatory effects in platelets57,67,68,75–77.
The chemical strategy that led to the identification of CFM-1571 was based on YC-1 as a lead structure53. The enzymatic assay used to detect putative sGC-activating derivatives was performed in the presence of submaximal concentrations of NO. Although not specified by the authors, this indicates that the screening approach was designed to identify compounds that synergize with NO, a clear characteristic of haem-dependent sGC stimulators.
The chemical structure of A-350619 (Abbott) and its analogues share no similarity with YC-155,78,79. However, like YC-1, A-350619 and its derivatives directly activate sGC and synergize with NO78. Oxidation of the sGC haem moiety by ODQ reduced the effectiveness of A-350619 in stimulating sGC. Although no competition-binding experiments have been published, Miller and co-workers proposed a common binding site for both structural classes78, based on findings showing that A-350619 and YC-1 did not have an additive effect on sGC activation when combined.
A rapid and highly sensitive cell-based assay for cGMP suitable for fully automated ultra-high-throughput screening was a key tool for further pharmacological breakthroughs. In this assay, which uses a Chinese hamster ovary cell line, intracellular cGMP production is linked to bioluminescence via an increase in Ca2+ influx through the olfactory cGMP-gated cation channel CNGA2, in the presence of aequorin, a Ca2+-sensitive luminescence indicator. This cell line was used to screen more than 900,000 compounds, leading to the identification of the primary hit BAY W 1449, an amino dicarboxylic acid80,81. A chemical derivatization programme was initiated and, in 2002, BAY 58–2667 was selected from a series of approximately 800 analogues as the first NO-independent activator of sGC that showed completely different characteristics to any of the known haem-dependent sGC stimulators. The activation of sGC by this compound was even stronger after oxidation or removal of the prosthetic haem group, indicating a previously unknown mechanism of enzyme activation80. Very recently, another compound with comparable characteristics, HMR-1766, was described, although the two chemical structures show no similarity82. These two compounds founded the novel class of NO- and haem-independent sGC activators (TABLE 2).
BAY 58–2667 directly activates sGC with EC50 and Kd values in the low nanomolar range, making this compound the most potent NO-independent sGC activator reported to date70,80. In contrast to the haem-dependent sGC stimulators, BAY 58–2667 produces an additive, not synergistic, effect when combined with NO donors. Oxidation or removal of the prosthetic haem group potentiates BAY 58–2667-induced enzyme activation80. This unique characteristic can be explained by a mechanistic model in which the sGC prosthetic haem is replaced by BAY 58–2667, thereby resulting in activation of the enzyme. BAY 58–2667 and the haem moiety compete via their negatively charged carboxylic groups for the unique sGC haem-binding motif Y-x-S-x-R17,18,83. This hypothesis, based on activity and binding assays as well as spectroscopic studies, is further strengthened by structural alignments showing that BAY 58–2667 is able to mimic the spatial structure of the sGC porphyrin ligand18. As only the oxidized haem can be effectively replaced with BAY 58–2667 (because of the strongly reduced affinity of the oxidized haem for the sGC haem-binding pocket in comparison to the reduced moiety), this compound is able to discriminate between both redox states18. Therefore, BAY 58–2667 activates haem-deficient or oxidized sGC by binding to the unoccupied haem-binding pocket or by replacing the weakly bound oxidized haem, whereas reduced sGC is virtually unresponsive to BAY 58–2667.
BAY 58–2667 relaxes blood vessels with a potency that is several orders of magnitude greater than the NO donors sodium nitroprusside (SNP) and 3-morpho-linosydnonimine84. The compound reduces coronary perfusion pressure in the rat Langendorff heart preparation and, like its predecessors BAY 41–2272 and BAY 41–8543, remains active in tissues made tolerant to glyceryl trinitrate80.
Submicromolar concentrations of HMR-1766 and the chemically related structure S-3448 (Sanofi-Aventis) directly activate purified sGC82. The addition of submaximal concentrations of NO results in an additive effect, whereas oxidation of the sGC haem moiety by ODQ strongly potentiates the HMR-1766 activating effect. Similarly to BAY 58–2667, competition between HMR-1766 and porphyrinic haemsite ligands was observed, indicating, at least in part, an overlapping binding site between this structural class and porphyrins. Schindler and co-workers proposed that HMR-1766 and S-3448 preferentially activate the oxidized haem-containing sGC, although activation of the haem-deficient enzyme was not ruled out82. In fact, the latter view was supported by the demonstration that a haem-free purified enzyme preparation could be activated by S-344882. Taken together these findings suggest that HMR-1766 and S-3448, similarly to BAY 58–2667, might be capable of activating both the haem-containing and the haem-deficient form of sGC. Future studies should clarify the exact mechanism of sGC activation by HMR-1766 and S-3448, and provide a direct comparison of these compounds with BAY 58–2667.
Arterial hypertension is one of the most important public health problems in the developed world. In 2000 in the US alone, there were at least 65 million adults with arterial hypertension, and its total prevalence rate was 31.3%85. Untreated, it leads to premature death by contributing to heart disease, stroke and renal failure. In 90–95% of cases the aetiology of arterial hypertension is unknown (essential hypertension), and therefore the therapy has to rely on symptomatic approaches rather than prevention or cure. A wide array of studies on human subjects suggests that alterations in NO–sGC–cGMP signalling that are unrelated to a decreased availability of the substrate for NO production are involved in its pathogenesis86–88. In arterial hypertension models (such as the spontaneously hypertensive rat (SHR)), endothelium-dependent vasodilation is impaired as a result, at least in part, of oxidative and nitrosative stress, and mRNA and protein levels of both the α1 and β1 subunits of sGC and sGC activity are reduced89–93. By contrast, NOS protein and activity levels seem to be upregulated to compensate for these changes30,93.
Clinically, the use of classical NO donors to treat arterial hypertension is problematic because of the development of tolerance following prolonged administration. Indeed, chronic exposure to endothelium-derived NO, as well as acute exposure to nitrovasodilator-derived NO, can lead to sGC desensitization without altering sGC expression94. Correspondingly, both acute cessation of endothelial NO formation by the pharmacological inhibition of NO synthase (NOS) in wild-type mice and chronic deficiency of NO in endothelial NOS−/− mice restore the sensitivity of sGC to NO and enhance vascular smooth muscle relaxation in response to nitrovasodilator agents94. There are therefore at least two mechanisms affecting sGC in vivo: downregulation of mRNA and protein levels, and desensitization. Non-NO-based sGC stimulators and activators potentially offer a major advantage over NO-releasing drugs by providing a means of sGC activation that is independent of NO–superoxide interactions, metabolic NO formation and NO-induced sGC desensitization (FIG. 1).
In normotensive animals, BAY 41–8543 lowers blood pressure and increases coronary blood flow67. Of note, a 500-fold higher dose of YC-1 is needed to achieve a comparable haemodynamic response50,54,95. In addition, reduction of blood pressure by YC-1 lasts only a few minutes95, whereas that of BAY 41–8543 is still present two hours after intravenous administration. In conscious SHR, orally applied YC-1 is devoid of any antihypertensive activity, unlike BAY 41–8543, which produces a dose-dependent and long-lasting (up to 24 hours) antihypertensive effect without causing tachyphylaxis following multiple administrations. In this experimental setting, BAY 41–8543 is about threefold more potent than BAY 41–227268. Furthermore, BAY 41–2272 and BAY 41–8543 prevent the increase in blood pressure induced by NOS inhibition in rats, attenuate cardiac remodelling and provide cardiorenal protection, which is associated with a reduction in mortality57,68,96. The pharmacological profiles of the sGC stimulators BAY 41–2172 and BAY 41–8543 therefore suggest that this class of compounds could have unique clinical utility in the treatment of essential hypertension and associated cardiac and renal complications.
The sGC activator BAY 58–2667 is the first NO-independent compound that has a haemodynamic profile similar to that of organic nitrates. It is about two orders of magnitude more potent as a vasodilator in isolated arterial and venous rings than BAY 41–227254,57. In anaesthetized rats, intravenous administration of BAY 58–2667 produces dose-dependent and long-lasting hypotension. ODQ does not inhibit the activity of BAY 58–2667 but, conversely, causes a significant increase in the potency and duration of the vasodilation46,97. These preliminary results suggest that, unlike glyceryl trinitrate, the efficacy of BAY 58–2667 is not reduced, and is in fact augmented, by oxidative stress, which is due to the targeting by this compound of the oxidation-impaired or haem-free sGC. This compound therefore represents an entirely new pharmacological principle, which selectively targets diseased blood vessels for antihypertensive and diagnostic purposes (FIG. 4).
Acute and chronic pulmonary hypertension (PH) are life-threatening conditions characterized by increased pulmonary arterial pressure that results from excessive pulmonary vasoconstriction. Chronic PH is also associated with pulmonary vascular remodelling, localized thrombosis and right heart hypertrophy. Although approaches for the diagnosis of PH have evolved dramatically in recent years, treatment of PH largely remains palliative98.
In the normal lung, the vascular endothelium and airway epithelium produce NO that regulates the processes required to match vascular perfusion with alveolar ventilation. Increased NO production causes a local increase in the synthesis of cGMP, which leads to pulmonary vasodilation. However, in PH the pulmonary production of NO and other endogenous vasodilators, such as prostacyclin, becomes markedly impaired, whereas vasoconstrictors, such as thromboxane A2 and endothelin, are released in excess98,99.
Inhalation of low concentrations of gaseous NO produces pulmonary vasodilation in well-ventilated lung regions without causing arterial hypotension. This selective pulmonary vasodilation is achieved because inhaled NO is rapidly bound by haemoglobin after entry into the intravascular space100. Limitations of NO inhalation as a chronic therapy for PH include the short duration of pulmonary vasodilation after NO is discontinued, the development of methaemoglobinaemia following inhalation of high concentrations of NO gas, and the observation that not all PH patients respond to inhaled NO100. The vasodilator effect of NO is mediated via cGMP-dependent mechanisms, and so it has been proposed that inhibition of the cGMP-metabolizing PDEs would dilate the pulmonary vasculature by increasing cGMP levels in pulmonary vascular smooth muscle cells. Subsequently, the PDE5 inhibitor sildenafil has been shown to produce pulmonary vasodilation in experimental models of PH, as well as in patients with chronic PH, and the drug has been approved for the treatment of this condition101–104. However, a significant proportion of patients with chronic PH fail to respond to sildenafil therapy, indicating that in these individuals endogenous NO production is reduced to such an extent that inhibition of cGMP degradation has no beneficial effects101. Accumulating experimental data suggest that under such conditions direct pharmacological activation of sGC could represent a promising therapeutic alternative.
In a sheep model of acute, chemically induced PH, intravenous infusion of BAY 41–2272 reduced mean pulmonary arterial pressure and pulmonary vascular resistance and increased transpulmonary cGMP release72. However, larger doses of BAY 41–2272 also elevated the cardiac index and produced systemic vaso dilation. Pharmacological inhibition of NOS abolished the systemic but not the pulmonary vasodilator effects of BAY 41–2272. The latter findings suggest that although BAY 41–2272 might act independently of endogenous NO production in the pulmonary vasculature, endogenous NO is required for the systemic vasodilator response to this compound. Furthermore, BAY 41–2272 potentiated and prolonged the pulmonary vasodilation induced by inhaled NO72. In a clinical setting, the capacity of an sGC stimulator to augment the efficacy of inhaled NO could result in an increased number of patients with PH responding to low concentrations of NO. In addition, the prolongation of the vasodilator effects of NO by an sGC stimulator might potentially facilitate chronic therapy with intermittently inhaled NO.
Similar haemodynamic effects were subsequently found following infusion of BAY 41–2272 in healthy ovine foetuses and a sheep model of persistent pulmonary hypertension of the newborn. Compared with sildenafil, the pulmonary vasodilator response to BAY 41–2272 was more prolonged105,106. In addition, BAY 41–2272 deceased mean pulmonary arterial pressure in dogs subjected to rapid ventricular pacing107. Moreover, in rodent models of chronic PH, treatment with BAY 41–2272 or BAY 58–2667 markedly attenuated an increase in right ventricular systolic pressure, right ventricular hypertrophy and structural remodelling of the lung vasculature108,109.
However, treatment of PH with oral or intravenous vasodilating agents, including NO-releasing drugs and stimulators or activators of sGC, can be associated with arterial hypotension and deterioration of arterial oxygenation due to pulmonary ventilation-perfusion mismatching72. Inhalation of vasodilating agents can provide targeted drug delivery to the lungs, thereby avoiding or reducing systemic side effects. Indeed, inhaling lipid–protein–sugar microparticles containing BAY 41–2272, BAY 41–8543 or BAY 58–2667 produced potent selective pulmonary vasodilation in lambs with acute PH without any adverse effects on pulmonary gas exchange110. Taken together, these preclinical studies provide strong evidence that direct pharmacological stimulators or activators of sGC, either alone or in combination with exogenous NO, could be an effective therapeutic intervention in PH, especially when endogenous NO–sGC–cGMP signalling is impaired. The sGC stimulator BAY 63–2521, which is structurally similar to BAY 41–2272 and BAY 41–8543, is now in clinical trials for the treatment of patients with chronic PH111.
Heart failure (HF), which is associated with a range of cardiovascular pathological conditions, is one of the major causes of hospitalization, morbidity and mortality worldwide, and represents a growing public-health issue. For example, in the US approximately 5 million people are affected by HF, more than 0.5 million new cases are diagnosed each year, and the total annual cost of managing patients with HF is approaching US$30 billion. Despite advances in treatment, the number of deaths from HF has increased steadily112. The progression of cardiovascular dysfunction to HF is complex and involves the activation of numerous secondary pathways113–115. In HF, sGC/pGC–cGMP signalling is disrupted either as a result of impaired production of NO or its excessive degradation, or chemical interactions with oxidants such as superoxide; in addition, inadequate release of the pGC ligands atrial and B-type natriuretic peptides, or the release of abnormal forms of these proteins, contributes to HF114,116. Conventional organic nitrates that activate sGC after bioconversion to NO have been used for more than a century to treat congestive HF; however, their therapeutic efficacy is limited because of the development of tolerance following chronic administration. The tolerance mechanism has been linked, in part, to increased vascular superoxide production, downregulation of mitochondrial aldehyde dehydrogenase, and supersensitivity to vasoconstrictors, secondary to a tonic activation of protein kinase C2,4.
In a canine model of congestive HF, intravenous administration of BAY 41–2272 increased cardiac output and renal blood flow and reduced the mean arterial, pulmonary arterial and pulmonary capillary wedge pressures, without decreasing the glomerular filtration rate or affecting the renin–angiotensin–aldosterone system, which is involved in the progression of the disease and development of complications107. Importantly, in this model BAY 41–2272 acted as a pure arterial vasodilator and, unlike glyceryl trinitrate, did not significantly decrease right atrial pressure. A possible explanation for these diverse effects is that sGC activation by glyceryl trinitrate depends on its biotransformation, which has been reported to be higher in the venous blood vessels117. Furthermore, the cGMP-independent actions of NO might also account for these differences.
Intravenous infusion of BAY 58–2667 or glyceryl trinitrate in anaesthetized dogs that were under autonomic blockade also produced a decrease in arterial blood pressure and reductions in diastolic pulmonary artery pressure and right atrial pressure. The duration of vasodilator effect was much longer for BAY 58–2667 compared with glyceryl trinitrate80. BAY 58–2667 is therefore the first new non-NO-based compound that has similar haemodynamic effects to glyceryl trinitrate, causing in vivo dilation of both arterial and venous blood vessels. In a canine model of congestive HF, intravenous administration of BAY 58–2667 resulted in dose-dependent reductions in cardiac preload and afterload, and a concomitant increase in cardiac output and renal blood flow without further neurohumoral activation118. BAY 58–2667 is currently undergoing clinical studies in patients with acute HF111. These findings suggest a novel mechanism for modulating the NO–sGC–cGMP pathway in the treatment of cardiac disease and for counteracting endothelial dysfunction without increasing vascular superoxide production and development of tolerance.
Atherosclerosis is the underlying cause of the majority of common cardiovascular diseases, which are the leading causes of death in Western societies119. The pathophysiology of atherosclerosis is complex and involves endothelial injury, which results in the accumulation of lipids and their uptake by monocytes. This is followed by platelet and monocyte adhesion and aggregation at the site of injury, and the release of factors that promote smooth-muscle proliferation and migration, which leads to the synthesis and deposition of extracellular matrix120,121. Percutaneous transluminal coronary and peripheral artery angioplasties have become widely available and effective treatments for patients with coronary and peripheral artery diseases. Although the incidence of restenosis has been reduced to 20–30% by the use of modern stents, which have further revolutionized angioplasty procedures, it remains a significant clinical problem with a limited number of therapeutic options122.
Nitric oxide is an important anti-atherosclerotic autocoid with anti-aggregatory effects on platelets and antiproliferative and dilatory effects on vasculature (FIG. 4). Local transfer of genes encoding NOS123,124 and local or systemic administration of NO donors125,126 attenuates neointima formation after experimental vascular balloon injury (a model of restenosis). Similarly, local adenovirus-mediated gene transfer of sGC α1 and β1 subunits partially restores sGC function and NO responsiveness in balloon-injured rat carotid arteries, resulting in reduced neointima formation in the presence of low concentrations of an NO donor31. There is recent evidence demonstrating that atherosclerosis is not only associated with decreased NO bioavailability, but also with alterations in signal-transduction components downstream of NO including, among others, sGC, particularly in neointima30. sGC might therefore represent an attractive, novel pharmacological target in the treatment of atherosclerosis and restenosis.
Indeed, topical application of the sGC stimulator YC-1 inhibited vascular smooth-muscle cell proliferation through cGMP-dependent mechanisms77,127 and also reduced the expression of transforming growth factor-β (TGFβ), focal adhesion kinase and matrix metalloproteinases (MMP2 and MMP9)77,128. YC-1 also markedly inhibited neointima formation after balloon-induced carotid artery injury in rats128,129. Likewise, in the same experimental model, BAY 41–2272 reduced the neointimal response to injury through antiproliferative and antimigratory actions on vascular smooth-muscle cells130. In addition to antiproliferative effects in vascular smooth-muscle cells, BAY 41–2272, BAY 41–8543 and BAY 58–2667 also inhibited platelet aggregation and thrombosis in various experimental models57,67,68,76,80. Moreover, BAY 41–2272 inhibited P-selectin expression on platelets and endothelial cells in vitro and reduced leukocyte rolling and adhesion in vivo, indicating a previously uncharacterized role for sGC in modulating the inflammatory response131. Collectively, the studies discussed above indicate that sGC stimulators might provide considerable therapeutic benefits in atherosclerosis and restenosis, as well as other inflammatory cardiovascular disorders.
Erectile dysfunction (ED) is the most common sexual dysfunction in men, affecting as many as 15–30 million people in the US alone132. The physiological process of erection is governed by a complex interplay between sympathetic, parasympathetic and nitrergic nerves, neurotransmitters, blood vessels and cavernous muscles133,134. It is well established that NO, which is synthesized during sexual stimulation in the nitrergic nerve terminals of the penis and also by the endothelial cells of blood vessels in the corpora cavernosa, is pivotal in the control of erectile function through its activating effects on the sGC–cGMP axis, leading to vascular smooth-muscle relaxation and penile erection135,136. The importance of this pathway is also supported by the inhibitory effect of both sGC and NOS inhibitors on NO-mediated relaxation of the corpus cavernosum, and by the relaxant effects of NO donors and PDE5 inhibitors on the corpus cavernosum133,134. Indeed, the NO-dependent increase in intra-cellular cGMP following PDE5 inhibition by sildenafil has been shown to be an effective approach in the treatment of ED137. However, nearly 30% of patients with ED do not respond to the PDE5 inhibitor therapy, implying that endogenous NO production is impaired to such an extent that inhibition of cGMP degradation provides no significant benefit. Under such conditions the use of direct, NO-independent stimulators of sGC might represent a promising alternative approach in the treatment of ED.
Studies with YC-1 have demonstrated that this compound is effective in relaxing the rat corpus cavernosum tissue in vitro and in enhancing erection induced by stimulation of the cavernous nerve and apomorphine in vivo78,138–140. The sGC stimulator A-350619 also relaxed isolated cavernosum tissue strips and induced penile erection in a conscious rat model78. Furthermore, the relaxation effect of both YC-1 and A-350619 on isolated cavernosum tissue was potentiated by the NO donor SNP78. BAY 41–2272 has also been shown to induce rabbit and human corpus cavernosum relaxation in vitro, and is ~30-fold more potent than YC-1141,142. Importantly, BAY 41–2272 enhanced nitrergic relaxations induced by electrical-field stimulation at concentrations known to have no PDE5 inhibitory activity69,142, further supporting the concept that BAY 41–2272 synergizes with NO. The effects of BAY 41–2272 and sildenafil on the anococcygeus muscle isolated from streptozotocin-induced diabetic rats, in which nitrergic relaxation responses are decreased, have been compared73. The residual nitrergic relaxation was enhanced by BAY 41–2272 but not by sildenafil. These results confirm the notion that endogenous NO derived from nitrergic nerves is significantly decreased in diabetes, and suggest that sGC stimulators could be more useful than PDE5 inhibitors in the treatment of diabetes-induced ED. Even when endogenous NO release was pharmacologically inhibited, BAY 41–2272 restored relaxant responses142. Moreover, BAY 41–2272 has been shown to cause penile erection in rabbits in vivo, producing a synergistic effect when combined with SNP143. The latter finding is very important because these new compounds are expected to enhance NO–sGC–cGMP signalling particularly during sexual stimulation, thereby facilitating natural penile erection. Taken together, this experimental evidence indicates that the use of sGC stimulators might represent a novel, promising strategy for the therapy of ED.
Renal fibrosis is the final common manifestation of a wide variety of chronic kidney diseases. Irrespective of the initial causes, progressive chronic kidney disease often results in widespread tissue scarring that leads to the complete destruction of kidney parenchyma and end-stage renal failure, a devastating condition that requires dialysis or kidney replacement144. There is a need for novel pharmacological approaches to delay or even prevent the onset of chronic kidney disease and fibrosis. Evidence suggests that the most relevant pathophysiological events that cause end-stage renal failure involve increased production of extracellular matrix by mesangial cells secondary to elevated expression of the profibrotic cytokine TGFβ, which is driven in a large part by increased angiotensin signalling144.
Based on the previously reported efficacy of NO donors in preventing matrix accumulation and tissue injury145, recent experimental studies evaluated whether an elevation in intracellular cGMP by direct stimulation of sGC would ameliorate renal disease. Administering BAY 41–2272 to rats with an acute form of glomerulonephritis attenuated renal dysfunction, as determined by the presence of proteinuria, an effect that correlated with decreased TGFβ production, matrix deposition and macrophage infiltration146. A subsequent study demonstrated that BAY 41–2272 elevated cGMP levels in mesangial cells, thereby reducing their proliferation and matrix production147. Interestingly, the disease process itself upregulates sGC protein expression with a concomitant increase in cGMP levels146,147, indicating that this effect could represent an endogenous protective mechanism, which adds further weight to the validity of utilizing pharmacological sGC stimulators to prevent renal disease.
Subsequent investigations in a chronic model of glomerulonephritis confirmed that BAY 41–2272 protects the kidney from progressive sclerosis and matrix deposition by limiting TGFβ expression148,149. The protective effect achieved by elevating cGMP via direct sGC stimulation with BAY 41–2272 was far superior to that produced by preventing degradation of cGMP using the PDE inhibitor pentoxifylline149. Furthermore, in rats with subtotal nephrectomy (a model of chronic renal failure), treatment with BAY 58–2667 for 18 weeks lowered blood pressure, reduced left ventricular hypertrophy and cardiac arterial wall thickness, and slowed the progression of renal disease150. In summary, these preclinical studies suggest that sGC stimulation or activation could be useful for the management of patients with chronic renal disease.
Cirrhosis, a pathological condition defined by deranged hepatic architecture, is the final common pathway of nearly all chronic diseases of the liver. It is now recognized that hepatic stellate cells are primarily responsible for hepatic fibrosis and subsequent progression to cirrhosis. Although substantial progress has been made in understanding the pathogenesis of hepatic fibrosis during the past 20 years, the success of current pharmacological therapies is very limited151.
The NO-dependent activation of sGC followed by an increase in intracellular cGMP has been shown to reduce hepatic stellate cell activation152. BAY 60–2770, a close chemical analogue of BAY 58–2667, has been studied in rat models of liver fibrosis (a pig serum model, a carbon tetrachloride model and an accelerated model that combines both stimuli). Oral treatment with BAY 60–2770 in doses that do not affect systemic blood pressure prevented the increase in hepatic fibrous collagen and total collagen accumulation153. These preliminary data indicate that direct activation of sGC might provide a novel approach for the treatment of liver fibrosis of necro-inflammatory and immunological origin.
The NO-independent stimulators and activators of sGC represent one of the major innovations in drug discovery in recent years. The first group of these compounds comprises the haem-dependent sGC stimulators (including YC-1, BAY 41–2272, BAY 41–8543, A-350619 and CFM-1571). These compounds show a strong synergy with NO and a loss of activation after oxidation or removal of the prosthetic haem moiety of sGC. The second group comprises the sGC activators (including BAY 58–2667 and HMR-1766), which have been found to require neither NO nor haem, and demonstrate even more pronounced action on the oxidized form of sGC. Emerging preclinical evidence suggests that stimulators and activators of sGC might offer a significant advantage over and/or increase the efficacy of current therapies in modulating NO–sGC–cGMP signalling in various forms of cardiovascular, pulmonary, endothelial, renal, hepatic and sexual dysfunctions. Besides their potential use as therapeutics, these compounds might be utilized as vascular functional diagnostics to elucidate vascular oxidative stress and endothelial dysfunction. Further evaluation of these drugs and their alternative routes of administration in animal models of chronic diseases, such as chronic heart failure and chronic pulmonary hypertension, will be the next step of preclinical testing. In addition, the possibility of adverse effects has to be addressed. The sGC activators and the second-generation sGC stimulators are now entering clinical development. It remains to be established whether these two novel therapeutic modalities will fulfill their initial promise in the clinical arena.
The authors thank P. Sexton for generating the sGC homology model. This publication was supported in part by the National Heart, Lung, and Blood Institute and the Intramural Research Program of National Institutes of Health (USA) and the Alexander von Humboldt Foundation (Germany).
Competing interests statement
The authors declare competing financial interests: see Web version for details.
The following terms in this article are linked online to:
sGC | MMP2 | MMP9 | PDE5
International Conference on cGMP:
Access to this links box is available online.