cGMP modulation therapeutics for sickle cell disease

Short abstract

Sickle cell disease (SCD) is an inherited disease caused by the production of abnormal hemoglobin (Hb) S, whose deoxygenation-induced polymerization results in red blood cell (RBC) sickling and numerous pathophysiological consequences. SCD affects approximately 300,000 newborns worldwide each year and is associated with acute and chronic complications, including frequent painful vaso-occlusive episodes that often require hospitalization. Chronic intravascular hemolysis in SCD significantly reduces vascular nitric oxide (NO) bioavailability, consequently decreasing intracellular signaling via cyclic guanosine monophosphate (cGMP), in turn diminishing vasodilation and contributing to the inflammatory mechanisms that trigger vaso-occlusive processes. Oxidative stress may further reduce NO bioavailability in SCD and can oxidize the intracellular enzyme target of NO, soluble guanylate cyclase (sGC), rendering it inactive. Increasing intracellular cGMP-dependent signaling constitutes an important pharmacological therapeutic approach for SCD with a view to augmenting vasodilation, and reducing inflammatory mechanisms, as well as for increasing the production of anti-polymerizing fetal Hb in erythroid cells. Pharmacological agents under pre-clinical and clinical investigation for SCD include NO-based therapeutics to augment NO bioavailability, as well as heme-dependent sGC stimulators and heme-independent sGC activators that directly stimulate native and oxidized sGC, respectively, therefore bypassing the need for vascular NO delivery. Additionally, the phosphodiesterases (PDEs) that degrade intracellular cyclic nucleotides with specific cellular distributions are attractive drug targets for SCD; PDE9 is highly expressed in hematopoietic cells, making the use of PDE9 inhibitors, originally developed for use in neurological diseases, a potential approach that could rapidly amplify intracellular cGMP concentrations in a relatively tissue-specific manner.

Impact statement

Sickle cell disease (SCD) is one of the most common inherited diseases and is associated with a reduced life expectancy and acute and chronic complications, including frequent painful vaso-occlusive episodes that often require hospitalization. At present, treatment of SCD is limited to hematopoietic stem cell transplant, transfusion, and limited options for pharmacotherapy, based principally on hydroxyurea therapy. This review highlights the importance of intracellular cGMP-dependent signaling pathways in SCD pathophysiology; modulation of these pathways with soluble guanylate cyclase (sGC) stimulators or phosphodiesterase (PDE) inhibitors could potentially provide vasorelaxation and anti-inflammatory effects, as well as elevate levels of anti-sickling fetal hemoglobin.

Introduction

Sickle cell disease (SCD) constitutes a group of genetic disorders, caused primarily by a mutation in the hemoglobin subunit beta gene (HBB; c.20A>T; glutamic acid-valine; rs334), producing altered sickle hemoglobin, HbS.¹ Homozygosity for this mutation results in sickle cell anemia (SCA; HbSS), while compound heterozygosity for HbS in association with other hemoglobin variants or thalassemias results in SCD, where disease phenotype demonstrates similarities to that of SCA.^1,2 It has been estimated that approximately 300,000 children are born with SCD in the world every year, of which the great majority are in Africa.³

The pathophysiology of SCD is caused by the polymerization of deoxygenated HbS, which can disrupt the flexibility and architecture of the red blood cell (RBC), causing it to become sickle shaped. The rate and extent of HbS polymerization depend on the HbS concentration in the erythrocytes, pH, temperature, and local oxygen tension.^4,5 Alterations in the physical properties of the RBCs of SCD individuals can trigger the premature destruction of erythrocytes, leading to chronic hemolytic anemia, and induce a number of inflammatory pathways that culminate in the hemolytic and vaso-occlusive processes that characterize the pathophysiology of the disease.^1,6 SCD displays a range of severity, but in general it is associated with high morbidity and a decreased life expectancy, with a wide range of acute and chronic complications, including acute painful vaso-occlusive episodes (VOEs) that often require hospitalization, stroke, acute chest syndrome (ACS), pulmonary hypertension, autosplenectomy, retinopathy, nephropathy, and leg ulcers.^7–9

Present therapeutic options for SCD include cell-based therapies (hematopoietic stem cell transplantation and transfusion) and pharmacotherapy based largely on hydroxyurea therapy. At least 40 substances are currently in various stages of pre-clinical and clinical studies for the prevention or treatment of VOE in SCD, all of which have been developed based on the complex pathophysiology of the disease.^6,10 We summarize herein evidence for dysregulation of nitric oxide (NO)-cyclic guanosine monophosphate (cGMP) signaling in SCD and describe cGMP-modulating pharmacotherapeutics under investigation with a view to use in SCD.

Pathophysiology of SCD

Alterations in RBC physiology, caused by HbS polymerization, result in a vicious circle of constant intravascular hemolytic and vaso-occlusive processes, in association with a chronic inflammatory state.⁶ Sickle RBCs are less deformable and, therefore, rupture more easily in the circulation; it has been estimated that up to 10% of the total RBC blood volume may be destroyed every day in an individual with SCD and that approximately 30% of this hemolysis may occur intravascularly.^1,11 Hemolysis may have a huge impact on both NO biology and inflammatory processes in the vasculature. Upon intravascular RBC lysis, large amounts of cell-free hemoglobin (Hb) are released into the circulation.¹² When not compartmentalized inside the RBC, Hb is extremely reactive and can rapidly release heme. Free heme has potent inflammatory effects, and is able to activate toll-like-receptor-(TLR) mediated cell signaling,¹³ induces macrophage inflammasome formation,¹⁴ activates complement, and stimulates neutrophil extracellular trap release, amongst other effects.^15,16 In mice with SCD, heme infusion has been shown to induce TLR4-mediated endothelial activation and microvascular stasis, and trigger ACS.¹⁷ The proinflammatory effects of heme released during hemolysis, in in vivo situations, have though been questioned, since highly hydrophobic heme is rapidly bound to and neutralized by macromolecules, such as hemopexin or albumin, or lipids¹⁸; however, hemopexin levels are significantly depleted in SCD^19,20 and it is possible that heme may modulate inflammation in more complex models of sequential priming and activation processes.¹⁸

Vaso-occlusive processes are a defining characteristic of SCD, where they occur principally in the microcirculation, reducing tissue oxygenation and inducing tissue damage. Inflammatory cell activation leads to the production and secretion of molecules such as cytokines and chemokines, growth factors, eicosanoids, and peptides that propagate the inflammatory state and further activate other cells in the vasculature. Numerous cytokines, produced from multiple cell types, are elevated in steady-state SCD, including the leukocyte-derived cytokine, tumor necrosis factor (TNF)-α, which may be generated as an early consequence of ischemia-reperfusion²¹ and has potent effects on both leukocytes and endothelial cells. Vaso-occlusion is a multicellular process that occurs as a consequence of these altered inflammatory and molecular pathways; in vitro studies and studies of the microvasculature of SCD mice models indicate that vaso-occlusion is initiated by the adhesion of activated leukocytes,^22,23 RBCs,^24,25 and platelets to the endothelium²⁶ in a mechanism that is propagated by inflammatory processes. In turn, diminished blood flow may decrease local oxygenation and trigger RBC sickling, resulting in occlusion of the blood vessel. Decreased vascular NO bioavailability, as well as oxidative stress, endothelial dysfunction, and the expression of adhesive molecules on blood cells and on the endothelium^1,6 all drive the vaso-occlusive process, and therefore represent potential therapeutic targets in SCD.

Evidence for the downregulation of NO-cGMP-dependent signaling in SCD pathophysiology

NO is a free radical signaling gas produced by the NO synthases (NOS) during the conversion of arginine to citrulline.²⁷ Endothelial-derived NO, produced by endothelial NOS (eNOS), plays a pivotal role in vascular homeostasis,²⁸ diffusing across the endothelial cell membrane into the adjacent smooth muscle and binding to the ferrous (Fe²⁺) heme of soluble guanylate cyclase (sGC). Activated sGC converts guanosine triphosphate (GTP) into cyclic guanosine monophosphate (cGMP), a second messenger that activates cGMP-dependent protein kinases (PKG) to promote a cascade of effects, resulting in calcium removal from smooth muscle cells, in turn relaxing blood vessels and increasing blood flow via vasodilation.^29,30 In addition to promoting vasodilation, NO also maintains vascular homeostasis by inhibiting platelet activity, adhesion and aggregation^31–34 and decreasing leukocyte adhesion and function^34–37 in a cGMP-dependent fashion. NO also inhibits endothelial adhesion activation and adhesion molecule expression.^38,39

In SCD, and other hemolytic diseases, the bioavailability of NO is compromised (Figure 1), largely as a consequence of its reaction with cell-free Hb. Upon intravascular hemolysis, cell-free extracellular oxyHb reacts with NO, rapidly and irreversibly forming methemoglobin and nitrate.^12,40 In addition, arginase, also released from the RBC during hemolysis, depletes arginine, reducing substrate availability for NO synthesis by eNOS.⁴¹ NO can also be consumed by reactions with reactive oxygen species (ROS), such as the superoxide radical, which are generated as a result of ischemia and reperfusion processes, and HbS autooxidation amongst other mechanisms.^8,42,43 Furthermore, anti-oxidant defense mechanisms are impaired in SCD, due to reduced antioxidant enzyme and oxygen radical scavenger levels, also contributing to the consumption of NO by ROS.⁴⁴

Figure 1.

Box figure of mechanisms leading to dysregulated NO-cGMP-dependent signaling and associated clinical complications in SCD. NO: nitric oxide: RBC; red blood cell: sGC; soluble guanylate cyclase.

Evidence for altered NO biology, and therefore altered cGMP-dependent signaling, in SCD pathophysiology has existed for some time (Figure 1); modulations in plasma NO metabolite levels were reported, in 1995, in SCD patients during acute VOEs,⁴⁵ while inhibition of NO synthesis in rats infused with sickle RBC induced their adhesion to the cerebral microvasculature, resulting in vaso-occlusion.⁴⁶ Lower levels of NO metabolites were later shown to correlate with higher pain scores in SCD patients hospitalized for acute VOEs.⁴⁷ However, the primary role of hemolysis in reduced NO bioavailability in SCD and the potential consequences of this effect was first highlighted by Reiter et al. in 2002,¹² who showed that NO-dependent increases in forearm blood flow were inhibited proportionately to cell-free Hb concentrations in SCD individuals, indicating that the decompartmentalization of Hb into plasma diverts NO from homeostatic vascular function in these patients. The induction of acute intravascular hemolysis in an animal model was later shown to produce dose-dependent systemic vasoconstriction and impair renal function, secondary to the stoichiometric oxidation of endogenous NO by cell-free Hb.⁴⁸ The vasoconstriction caused by this acute hemolytic process could be attenuated by the inhalation of NO gas in this experimental model.⁴⁸

Consumption of NO by cell-free Hb and, therefore, dysregulation of cGMP-dependent signaling could have further vascular consequences. There is evidence to suggest that NO can inhibit HbS polymer formation in RBC by abolishing the excess positive charge of HbS and increasing its oxygen affinity; therefore, reduced NO could augment RBC sickling.⁴⁹ Exposure of mice with SCD to hypoxia reduces cGMP in the lungs of animals, in association with elevated xanthine oxidase (an enzyme that generates ROS), impaired eNOS function, and acute lung injury.⁵⁰ Interestingly, RBCs possess NOS and inhibition of this NOS activity by oxidative stress may diminish RBC deformability.⁵¹

There is also substantial evidence that reduced NO bioavailability may augment cell–cell interactions and vascular inflammatory processes in SCD. NO inhibits sickle RBC adhesion to the endothelium,⁵² and hypoxia and low NO bioavailability may synergistically augment sickle RBC adhesion to endothelium via upregulation of endothelial P-selectin.⁵³ Leukocytes present augmented adhesive properties in SCD and both pharmacological NO donation and sGC stimulation are reported to inhibit the adhesion of SCD neutrophils to ligands found on the vascular wall.⁵⁴ Furthermore, the induction of acute hemolytic processes in mice, to yield similar levels of plasma cell-free Hb/heme to those seen in mice with SCD, results in substantial systemic and vascular inflammation, leading to cellular recruitment mechanisms resembling those observed in the SCD mouse microcirculation.⁵⁵ This inflammation was found to be associated with the modulation of plasma NO metabolites and could be inhibited by the co-administration of NO donor agents in an apparently sGC-dependent manner.⁵⁵

Thus, reduced NO in SCD can culminate in an imbalance between vasodilation and vasoconstriction, leading to endothelial dysfunction, which in the case of SCD may contribute to vaso-occlusive processes and ultimately acute and chronic complications, such as pulmonary hypertension.⁹ Furthermore, in addition to facilitating vasodilation, NO-cGMP signaling pathways are important for preventing leukocyte recruitment and the adhesive interactions of both platelets and endothelial cells in blood vessels (Figure 1);⁴⁰ therefore, it would seem reasonable to conclude that modulation of levels of NO and/or its intracellular second messenger, cGMP, could represent an effective approach to reducing vaso-occlusive processes in SCD (Figure 2).

Figure 2.

Role of cGMP-dependent signaling and potential for cGMP-modulation therapies in SCD. Nitric oxide (NO) is generated from the conversion of L-arginine to L-citrulline by nitric oxide synthases (NOS), particularly endothelial NOS (eNOS). Once diffused across the endothelium, NO binds to the heme moiety of soluble guanylate cyclase (sGC) in smooth muscle cells and other cell types, promoting the conversion of guanosine triphosphate (GTP) to cyclic guanosine monophosphate (cGMP), which in turn activates cGMP-dependent protein kinases (PKG). The cascade of reactions triggered by this pathway plays a pivotal role in vascular homeostasis. Calcium removal from smooth muscle cells promotes relaxation of blood vessels and increases the blood flow. Besides promoting vasodilation, the cGMP-signaling pathway inhibits platelet activation and aggregation, endothelial activation, and adhesion molecule expression, decreasing adhesive interactions between platelets, erythrocytes, leukocytes and the endothelium. Additionally, sGC activation upregulates the expression of the gene encoding γ-globin (HBG) in erythroid cells, increasing fetal hemoglobin (HbF) levels in the erythrocytes and preventing HbS polymerization and hemolysis in sickle cell diseases (SCD). Since NO bioavailability is decreased in SCD due to hemolysis, NO inhalation, nitrite administration, or hydroxyurea are therapies that directly restore intravascular NO levels in varying efficacies, while supplementation with L-arginine provides the substrate for NO synthesis. sGC stimulators (heme-dependent) and activators (heme-independent) are a class of drugs that modulate sGC activity independently of NO. sGC activators increase the activity of sGC only in its oxidized Fe³⁺heme or heme-free (apo) state and are particularly efficient for activating sGC under oxidative stress conditions. Additionally, the inhibition of phosphodiesterases (PDEs), a group of enzymes that regulate cGMP by catalyzing its hydrolysis and degradation, has been investigated as therapy for SCD. Sildenafil is a selective inhibitor of PDE5, expressed in smooth muscle cells, platelets and corpus cavernosum, whereas IMR-687, BAY 73–6691 and PF04447943 inhibit PDE9, an isoform highly expressed in hematopoietic cells. HbS: hemoglobin S; VCAM: vascular cell adhesion molecule-1; ICAM: intercellular adhesion molecule-1. (A color version of this figure is available in the online journal.)

Role of cGMP-dependent signaling in fetal hemoglobin modulation

High levels of fetal hemoglobin (HbF) in the RBC are known to ameliorate the pathophysiology of SCD. HbF, formed by the combination of two α-globin and two γ-globin proteins, is normally synthesized during fetal life, where switching to the production of adult Hb, HbA (α2β2), or HbS (α2βS2) in the case of SCD individuals, occurs during the first year of life.^56,57 The levels of HbF are modulated genetically in SCD and those individuals with high levels of HbF generally display milder disease⁵⁷ as HbF cannot form part of the HbS polymer and therefore reduces Hb polymerization and red cell sickling. HbF switching to adult Hb is regulated by a complex mechanism of gene regulation, in which the zinc-finger transcriptional factor, BCL11A plays an important role in the repression of fetal Hb production.⁵⁸

However, despite the complex transcriptional regulation of γ-globin gene expression, evidence exists that cGMP-dependent signaling in erythroid cells can modulate this gene’s expression.^59,60 sGC activators or cGMP analogs were found to increase the expression of the γ-globin gene (HBG) in both erythroleukemic cells and primary erythroblasts from healthy subjects and patients with beta-thalassemia.⁵⁹ Moreover, HbF induction by hemin and butyrate could be abolished by inhibiting sGC or PKG, suggesting that the sGC-PKG pathway can regulate HBG expression.⁵⁹ sGC activity can be identified in RBCs and the enzyme apparently remains responsive to NO and sGC stimulation even in patients with endothelial dysfunction.^61,62 cGMP levels have been reported as significantly higher in RBC of patients with SCA than healthy individuals, and RBC cGMP levels correlated with fetal Hb (HbF) levels in SCA, but not with reticulocyte count, indicating that augmentation of cGMP levels by NO in erythroid cells may constitute a mechanism for induction of HbF.⁶¹ Further reports indicate that cross talk between cGMP and cyclic adenosine monophosphate (cAMP)-dependent pathways may also contribute to HbF regulation, where these signaling molecules may share a common induction pathway, with evidence that cAMP dependent signaling is associated with downregulation of BCL11A expression, indicating a mechanism by which these signaling pathways may repress β-globin gene expression.^63,64

NO donor and sGC activation properties of hydroxyurea

Hydroxyurea, also known as hydroxycarbamide, is the only drug approved by both the FDA and EMA for the prevention of recurrent painful vaso-occlusive crises in pediatric and adult patients suffering from symptomatic SCD. Pharmaceutical grade L-glutamine, an amino acid with antioxidant properties, was recently approved by the FDA for use in SCD, but hydroxyurea remains the mainstay for SCD pharmacotherapy, at this time. Hydroxyurea therapy significantly elevates HbF and reduces the incidence of acute VOEs, hospital admissions, ACS, and the need for blood transfusions in patients with SCD.^65–69 Hydroxyurea is generally well tolerated by patients, however, there are some concerns regarding the effects of the drug on spermatogenesis⁷⁰ and a recent Cochrane Review concluded that evidence of the long-term benefits of hydroxyurea for the prevention of the chronic complications of the disease is insufficient at present.⁷¹

Hydroxyurea is a cytostatic agent that is also found endogenously at varying concentrations in the plasma and tissues of humans and animals, possibly acting as a natural defense agent against infections.^72–74 This molecule has NO donor and sGC activating properties and can be metabolized and oxidized by free Hb and heme, resulting in NO production from the –NHOH portion of hydroxyurea (Figure 2).^75–78 Hydroxyurea also induces NO production by enhancing eNOS phosphorylation and activity in endothelial cells⁷⁹ and may react with peroxidases to rapidly form NO in the presence of hydrogen peroxide (H₂O₂).^80,81 Additionally, evidence suggests that hydroxyurea can interact directly with the Fe²⁺-heme of sGC, promoting the iron nitrosylation of sGC, and consequently activating cGMP production.⁸²

Hydroxyurea induces cGMP elevation and gamma-globin gene expression in K562 erythroleukemic cells and human erythroid progenitor cells in a mechanism that can be abolished by the inhibition of sGC.⁶⁰ It has also been suggested that these cGMP-dependent molecular mechanisms may also be involved in the cytostatic effects of hydroxyurea in erythroid progenitor cells.⁸² Furthermore, hydroxyurea administration increases plasma NO metabolites and cGMP levels in SCA patients (in steady state and experiencing VOE), in a temporal manner.^77,83 Elevated RBC cGMP in hydroxyurea-treated SCA patients has been shown to correlate with individual HbF levels,⁶¹ although cAMP production is also reportedly required for full induction of HbF by hydroxyurea in human CD34(+) erythroid cell cultures (Table 1).⁶³

Table 1.

cGMP modulating drugs and drug candidates for SCD therapy.

Classification of agents	Mechanism of action	Agents	Pre-clinical evidence in SCD	Clinical evidence in SCD	Limitations
Nitric oxide donors	Improvement in NO bioavailability.	NO inhalation	Improvement in survival and lung injury following exposure to hypoxia and hypoxia-reperfusion in SCD animal model84,85	Improvement in pain scores during VOC86,87	No effects on Hb oxygen affinity; failure to improve outcomes of VOC compared with placebo; no efficacy for the treatment of ACS84,85,86
		L-arginine supplementation	Increase of NO metabolite levels; reduction of lipid peroxidation; increase of antioxidant levels in SCD animal model88	Benefits on pulmonary hypertension in SCD; amplification of NO response when co-administered with HU; reduction in pain scores and opioid use in children experiencing VOC89–91	Failure to demonstrate long-term clinical efficacy; alterations in redox potential of RBCs; no effects on hospital length of stay during VOC91,92
		Hydroxyurea	Evidence for cGMP-dependent effects:Anti-inflammatory effects via an NO-sGC dependent mechanism in SCD animal model.55 Induction of γ-globin expression in human erythroid progenitor cells via a sGC-dependent pathway60	Evidence for cGMP-dependent effects: Association of RBC HbF with intracellular [cGMP], indicating a role for sGC-mediated elevation of HbF; increase of plasma NO metabolites and cGMP levels in patients on HU61,77,83	Cytotoxic effects and long-term benefits for chronic organ damage still unclear70,71
		Nitrite	Improvement in SCD RBC deformability in vitro; reductions in RBC, leukocyte and platelet adhesion; reduction in hemolysis rate in vitro and in SCD mice93	Increase in regional blood flow in patients with SCD in steady state94	Reduced vasodilatory sensitivity in SCD individuals, compared with those without SCD94
sGC stimulators	Elevation of cGMP by sGC in a heme-dependent manner.	BAY 41-2272	Reduction of SCD neutrophil adhesion in vitro; improvement of cavernosal relaxation in SCD animal model95,96	No clinical studies for SCD.	Unsuitable for clinical use: Low metabolic stability; low oral bioavailability97
		Riociguat (BAY 63-2521)	No pre-clinical studies for SCD.	Safe and well tolerated, in SCD patients with chronic thromboembolic pulmonary hypertension: Significant improvements in exercise capacity, NT-proBNP, and RVSP in some of the treated patients98	Short half-life in humans and requires thrice-daily dosing;99 potential hypotensive effects should be monitored in SCD.
		Vericiguat (BAY 102-1189)	No pre-clinical studies for SCD.	No clinical studies for SCD.	Not applicable.
		Olinciguat (IW-1701)	Induces in vitro expression of γ-globin gene in erythroleukemic. Cells decreases leukocyte recruitment in C57BL/6 mice following an inflammatory stimulus100,101	Phase 2 double-blind, placebo-controlled multi-site trial for use in SCD currently undergoing (NCT03285178). Results not yet available.	Not yet reported.
sGC activators	Production of cGMP by increasing oxidized sGC activity in a heme-independent manner.	BAY 54-6544	Decreases cardiac remodeling and increases vaso-relaxation in SCD animal model102	No clinical studies for SCD.	Not applicable.
		BAY 60-2770	Decreases adhesive properties of human SCD neutrophils in vitro; decreases leukocyte recruitment and vaso-occlusive processes in SCD animal model103	No clinical studies for SCD.	Not applicable.
Phosphodiesterase inhibitors	Prevention of intracellular cGMP degradation.	Sildenafil	Beneficial effects on priapism in SCD animal models104	Prevention and resolution of recurrent ischemic episodes of priapism in SCD105,106A clinical trial (NCT00492531) to evaluate the effects of sildenafil in subjects with SCD with high TRV was terminated early due to increased hospitalizations for pain in patients on sildenafil; no favorable effects on the evaluated parameters were observed107	Risk of augmented pain processing107
		BAY 73-6691	Decrease of in vitro adhesive properties of SCD neutrophils. Inhibition of leukocyte recruitment, vaso-occlusive processes and survival in SCD animal model95,108,109	No clinical studies in humans.	Not applicable.
			Reduction of leukocyte-platelet aggregates and endothelial activation in SCD animal model110	A phase 1 clinical trial (NCT02114203) has been conducted to assess the safety, tolerability, pharmacokinetics and pharmacodynamics in subjects with SCD. Soluble E-selectin and heterocellular aggregates reportedly decreased in patients on PF-04447943.111 Apparent overall safety observed, although adverse events were observed in 3 out of 22 patients on PF-04447943 (vaso-occlusive crisis, biliary colic and pneumonia).	Not yet reported.
		IMR-687	Inhibition of microvascular stasis following hypoxia; increase in RBC HbF content; decrease in RBC sickling and leukocytosis in SCD animal model112	A randomized, placebo-controlled, multicenter study phase 2 trial (NCT03401112) to determine the safety, pharmacokinetics, and preliminary pharmacodynamics of escalating doses of IMR-687 in patients with SCD is underway. Results not yet available.	Not yet reported.

HU; hydroxyurea: NO; nitric oxide; NT-proBNP; N-terminal pro-brain natriuretic peptide: RBC; red blood cell: RVSP; right ventricular systolic pressure: SCD; sickle cell disease: TRV; ricuspid regurgitant velocity.

Hydroxyurea also exerts acute and immediate anti-inflammatory effects via an NO-sGC dependent mechanism in animal models.⁵⁵ The induction of acute inflammatory responses to hemolysis can be abolished by a single administration of hydroxyurea in a reaction that is dependent upon NO generation and sGC activity.⁵⁵ As such, in addition to evidence that hydroxyurea may mediate its HbF-elevating effects via an NO-sGC dependent mechanism, data provide perspectives for the use of hydroxyurea as an acute treatment for SCD and other hemolytic diseases by mechanisms that are independent of HbF induction, and probably occur via generation of intravascular NO and sGC activation.

NO-based therapies in SCD

Having defined a critical role for reduced NO bioavailability in SCD, therapeutic strategies that improve NO bioavailability have been extensively studied (Table 1),⁴⁰ with a view to enhancing smooth muscle relaxation, vasodilation, and increasing regional blood flow. Use of inhaled NO was first hailed as a prospective therapy for SCD, with potential for decreasing arterial NO consumption by oxidizing and nitrosylating cell-free Hb, and therefore restoring endogenous NO availability.¹² In a mouse model of SCD, NO inhalation ameliorated survival and lung injury following exposure to hypoxia and hypoxia/reperfusion, respectively.^84,85 Inhaled NO gas demonstrated safety and some therapeutic effects in small studies of SCD patients suffering from acute vaso-occlusive crisis (VOC), with improvements seen in pain scores.^86,87 However, no effects of NO inhalation on the oxygen affinity of Hb have been found¹¹³ and, in a larger placebo-controlled study in 150 SCD patients hospitalized with VOC, the use of inhaled NO, compared with placebo, did not improve time to crisis resolution, length of hospitalization, visual analog pain scale scores, cumulative opioid usage, or the rate of ACS.¹¹⁴ While increased nitrate levels were observed following NO exposure in this study and indicated systemic effects of NO, levels of blood nitrite, which may exert important benefits at ischemic sites, were not significantly increased.¹¹⁴ Furthermore, when used for the treatment of ACS in SCD, inhaled NO did not ameliorate treatment failure rate.¹¹⁵

Low levels of L-arginine, the substrate for NO synthesis, are observed in SCD patients, particularly during VOC, which could reflect a state of depletion that results in decreased NO production.^47,116 Supplementation with oral L-arginine has, therefore, been investigated for efficacy in SCD, as reduced L-Arginine levels appear to be rate-limiting for NO production during VOC.¹¹⁷ In mouse models of SCD, dietary arginine supplementation significantly increased NO metabolite levels, reduced lipid peroxidation and elevated antioxidant levels,⁸⁸ supporting the rationale for use of L-arginine in human SCD. L-Argine therapy has shown benefits for pulmonary hypertension in SCD, with reduced estimated pulmonary artery systolic pressure reported after five days of therapy in 10 SCD patients with pulmonary hypertension.⁸⁹ While L-arginine administration appears not to significantly augment NO metabolite production in SCD patients at steady state, data suggest that co-administration of L-arginine with hydroxyurea may amplify the NO response in SCD at steady state.⁹⁰ Use of L-arginine for 12 weeks in a small number of SCD patients in steady-state on hydroxyurea, however, failed to demonstrate clinically detectable efficacy and actually induced alterations in redox potential in red cells,⁹² with data indicating that the L-arginine administered may be diverted to ornithine production due to the elevated levels of plasma L-arginase. L-arginine may, though, display greater benefits in SCD patients experiencing acute vaso-occlusive pain; in a double-blinded placebo-controlled trial, 38 children with SCD that were hospitalized for pain were randomized to receive L-arginine or placebo for five days or until discharge. L-arginine was found to be safe to use and, importantly, a very significant reduction in total parenteral opioid use and lower pain scores at discharge were observed in those that received L-arginine, compared to placebo, although there was no significant difference in the hospital length of stay.⁹¹

Given the disappointing lack of efficacy of the use of inhaled NO in SCD, use of nitrite as therapy has been suggested. The nitrite anion (NO²⁻) is a potent and fast vasodilator at near-physiological concentrations and acts as an endocrine reservoir of NO; furthermore, it does not induce tolerance, as observed with the organic nitrates.¹¹⁸ Importantly, nitrite can be reduced to NO by deoxygenated Hb,^119,120 signifying that it can augment vasodilation under hypoxic conditions.¹²¹ Due to the constant processes of ischemia/reperfusion that occur in the vasculature² and the fact that nitrites can be bioactivated in the presence of RBCs,¹²² these agents make very attractive candidates for use in SCD, although the rate of nitrite reduction by HbS in polymer form may be decreased, when compared to HbA and non-polymerized HbS.¹²³

In vitro studies and in vivo studies with transgenic SCD mice have shown that nitrite may also improve RBC deformability as well as reduce RBC, leukocyte, and platelet adhesion, in addition to reducing hemolysis.⁹³ When administered to a small number of individuals with SCD and in steady state, sodium nitrite infusions augmented plasma nitrite concentrations and augmented forearm blood flow without causing hypotension or clinically significant methemoglobinemia.⁹⁴

Potential for sGC agonist therapy in SCD

While decreased NO signaling may play a major role in SCD pathophysiology, restoration of NO-dependent signaling more directly with NO inhalation and L-arginine supplementation in patients has shown limited efficacy, perhaps due to the fact that the mechanisms that limit NO production and bioavailability are sustained even during the therapy. Furthermore, the effects of NO donors such as nitrates in cardiovascular disease, for example, are limited by increased oxidative stress and tolerance.¹²⁴ As such, approaches that aim to directly stimulate the molecular target of NO could provide a more efficient approach to increasing the downstream effects of NO signaling in SCD. The NO receptor, sGC, is the target of two novel classes of drugs that enhance cGMP production and signaling. These compounds, denominated as the sGC stimulators and sGC activators, boost the enzymatic activity of sGC to generate cGMP, independently of the presence of NO (Figure 2 and Table 1).¹²⁵

sGC stimulators

The sGC stimulators (NO-independent heme-dependent sGC stimulators) were the first drugs to be developed to specifically modulate sGC activity; this class of drugs acts on the native conformation of the sGC enzyme in which the heme moiety is maintained. The sGC stimulators have a dual mode of action, as they synergize with endogenously available NO, and are also capable of directly binding and stimulating native sGC, to produce cGMP, independently of NO.^125,126 YC-1 was the first heme-dependent sGC stimulator to be characterized and was found to inhibit platelet aggregation independently of the presence of NO by stimulation of cGMP synthesis^127,128; however, while this compound presents low substrate specificity and a poor pharmacokinetic profile, it paved the way for the development of more potent and specific sGC modulators.¹²⁵ Given the antiproliferative, antifibrotic, antiinflammatory, proapoptotic, and neuroprotective effects of cGMP, sGC stimulators potentially have a wide range of beneficial effects.⁹⁷

With regard to effects specifically in SCD, the potent sGC stimulator, BAY 41–2272, has been shown to reduce the increased adhesive properties of neutrophils from SCD patients,⁹⁵ while in mice with SCD, this compound improves cavernosal relaxation, indicating potential for this class of drugs for preventing leukocyte recruitment (and therefore vaso-occlusive processes) and priapism in SCD.⁹⁶ BAY 41–2272, however, has low metabolic stability and oral bioavailability, making it unsuitable for clinical use.⁹⁷ Riociguat (BAY 63–2521), a compound with enhanced pharmacokinetics and oral bioavailability, was the first sGC stimulator licensed for clinical use and was approved in 2013 for the treatment of pulmonary hypertension.¹²⁹ Use of riociguat in a small case series of SCD patients with chronic thromboembolic pulmonary hypertension was recently reported; riociguat therapy was found to be safe overall and well tolerated, and showed clinical efficacy in some of the patients evaluated.⁹⁸ Riociguat, however, has a short-half life in humans and requires thrice-daily dosing.⁹⁹ Other sGC stimulators in clinical development, include vericiguat (BAY 102–1189), which has a longer half-life than riociguat and is in development for the treatment of chronic heart failure¹³⁰ and olinciguat (IW-1701), which is currently in clinical development specifically for use in SCD, and also achalasia. Olinciguat is an orally available sGC stimulator, with potential anti-inflammatory and vasodilating properties that has been shown to elevate γ-globin mRNA expression in erythroleukemic K562 cells.¹⁰⁰ In an in vivo study in C57BL/6 mice, prophylactic treatment with olinciguat decreased TNFα-stimulated adhesive interactions between leukocytes and endothelial cells in mice, and this effect was potentiated when olinciguat was combined with hydroxyurea.¹⁰¹ In a Phase 1b placebo-controlled, randomized, multiple-ascending-dose study in healthy subjects, olinciguat given once-daily for up to 14 consecutive days was well tolerated and no serious adverse events were observed. Olinciguat displayed rapid absorption, an adequate half-life for once-daily dosing and achieved plasma cGMP elevation.¹³¹ Olinciguat recently received orphan-drug status from the FDA as a potential treatment for SCD¹³² and a phase 2 double-blind, placebo-controlled multi-site trial (NCT03285178; STRONG-SCD) with a planned enrolment of 88 SCD patients is underway to evaluate the safety and tolerability, pharmacokinetics, and pharmacodynamics of three dose levels of olinciguat in steady-state individuals with SCD (aged 16–70 years), when administered for 12 weeks. Study completion is estimated for the first half of 2019. Crucially, based on their pharmacological mode of action, heme-dependent sGC stimulators may hold potential for use in combination therapy to amplify the NO-donating effects of hydroxyurea.

sGC activators

In contrast to the sGC stimulators, the sGC activators (NO-independent heme-independent sGC activators) increase the activity of sGC only in its oxidized Fe³⁺ or heme-free inactive (apo-) state.^133,134 sGC activators, therefore, trigger sGC activity independently of NO, but their efficacy appears to be additive to the effects of endogenous NO.^124,135 Since the sGC enzyme must be oxidized for these agents to have effect, they are particularly valuable for use in conditions of oxidative stress, making them extremely attractive for investigation in SCD, a disease characterized by oxidative stress.¹³⁶

Cinaciguat (BAY 58–2667) is a potent sGC activator that has been shown to promote vasodilation and attenuate pulmonary hypertension, endothelial dysfunction, platelet aggregation and thrombosis in experimental models.^134,137 Cinaciguat, however, has a short half-life and despite success in preclinical studies did not show significant efficacy in clinical studies of patients with acute decompensated heart failure, and in fact was associated with hypotension in some patients.¹³⁸ Another NO- and heme-independent sGC activator, ataciguat (HMR1766), improves vasomotor function and reduces platelet activation in rats with congestive heart failure,¹³⁹ but data regarding its clinical efficacy in clinical trials for aortic valve stenosis (NCT02049203), and neuropathic pain (NCT00799656) have not been published.

BAY60–2770 is another sGC activator with cardioprotective effects.¹⁴⁰ Preliminary data indicate that this sGC activator can abrogate the adhesive properties of neutrophils from SCD individuals and significantly decrease leukocyte recruitment and vaso-occlusive-like processes, in a mouse model of inflammatory SCD vaso-occlusion.¹⁰³ Interestingly, BAY 60-2770 is also able to provide renal protection in a mouse model of albuminuria,¹⁴¹ a property that could be important for protecting against the nephropathy that can occur in individuals with SCD.¹⁴² Like cinaciguat, BAY60-2770 is a heme-mimicking protein, and can stably insert into the sGC enzyme during protein biosynthesis and maturation.¹⁴³ Insertion of BAY60-2770 displaces heme in sGC, independently of its redox state, possibly altering the pharmacodynamic profile of the drug and may explain why sGC activators can induce hypotension when long infusion times are used.¹⁴³

With regard to the potential use of sGC activators in SCD, the chronic oral administration of the sGC activator, BAY 54–6544, was found to decrease cardiac remodeling in mice with SCD more efficiently than the sGC stimulator, BAY 41–8543, without altering systemic blood pressure.¹⁰² Furthermore, the BAY 54–6544 sGC activator improved ex vivo endothelium-dependent and -independent relaxation of the pulmonary artery of SCD mice, while the sGC stimulator was unable to augment vasorelaxation.¹⁰²

Data suggest that sGC is oxidized in the pulmonary arteries of SCD mice, hindering NO-induced responses¹⁴⁴; sGC activation may therefore represent a potential therapy that bypasses the need for NO-induced responses to improve vasorelaxation in SCD and potentially provide therapy for pulmonary arterial hypertension and cardiac remodeling in these patients. While clinical findings for cinaciguat were disappointing for treating acute decompensated heart failure, the compound did exert some clinical benefits, such as a rapid and sustained reduction in pulmonary capillary wedge pressure.¹³⁸ Therefore, investigations continue to identify sGC activators with improved pharmacokinetics that do not significantly alter systemic blood pressure, particularly for those pathologies in which oxidative stress is associated.

Phosphodiesterase inhibitor therapy in SCD

Phosphodiesterases (PDEs) are a class of enzymes that hydrolyze cGMP and cAMP, in turn diminishing intracellular cGMP and/or cAMP activity.¹⁴⁵ The PDE enzymes are divided into 11 families, each containing different isoforms. As each PDE has a different and specific tissue/cellular distribution and expression, they make ideal targets for drug development due to the potential for fewer adverse effects and greater specificity; furthermore, the low intracellular concentrations of PDEs make them good substrates for competitive inhibitors and their inhibition provides a very rapid manner of effectively amplifying intracellular cGMP levels.¹⁴⁶

PDE5 specifically hydrolyzes cGMP and is an important regulator of vascular smooth muscle contraction, especially in the penis and in the lung, where it is reportedly up-regulated in conditions of pulmonary hypertension.^147,148 Sildenafil is a selective PDE5 inhibitor that improves pulmonary hemodynamics and functional capacity in pulmonary hypertension¹⁴⁹ and preliminary investigations indicated that this compound could be beneficial in SCD pulmonary hypertension (Table 1).¹⁵⁰ However, a clinical trial (NCT00492531) conducted to evaluate the effects of sildenafil in subjects with SCD with high tricuspid regurgitant velocity (TRV) and decreased exercise capacity had to be terminated early due to an increased rate of hospitalizations for severe pain episodes in patients on sildenafil therapy compared with placebo,¹⁰⁷ and no favorable effects on the evaluated parameters were observed. In contrast, sildenafil has shown benefits for preventing and resolving recurrent ischemic episodes of priapism in SCD.^104–106

The PDE9A (PDE9) enzyme has the highest affinity of all the PDEs for cGMP.¹⁵¹ In contrast to PDE5, the expression of PDE9A is largely restricted to the brain,¹⁵¹ although it is also very highly expressed in hematopoietic cells, with even higher expression in the neutrophils and reticulocytes of patients with SCD.¹⁰⁸ PDE9 inhibitors have been largely developed for use in Alzheimer’s disease; however, there is evidence that PDE9 inhibition could be of benefit in SCD. BAY73-6691 is a potent and selective PDE9 inhibitor (PDE9i)¹⁵² that has been demonstrated to decrease the in vitro adhesive properties of SCD neutrophils.^95,108 Moreover, BAY73-6691 was found to elevate gene expression of γ-globin in erythroleukemic cells in vitro.¹⁰⁸ In vivo studies, in an inflammatory SCD mouse model, later found that inhibition of PDE9 with BAY73-6691 significantly inhibited leukocyte recruitment and vaso-occlusive processes in the cremaster microcirculation when given in a single intravenous administration. Furthermore, this PDE9i presented synergistic effects when administered intravenously together with hydroxyurea, reducing endothelial activation and augmenting animal survival in a cGMP-dependent manner.¹⁰⁹ Due to its high expression in hematopoietic cells, PDE9, therefore, represents a tissue-specific target to rapidly enhance cGMP activity, amplifying the NO-mediated acute beneficial effects of hydroxyurea in the leukocytes and endothelium, in turn inhibiting SCD vaso-occlusive processes.¹⁰⁹ Thus, the use of PDE9 inhibitors represents an extremely attractive approach in SCD, due not only to their immediate and acute anti-inflammatory effects in the vasculature and their relative hematopoietic tissue selectivity, but also their potential for elevating HbF synthesis when used chronically (Table 1).¹⁰⁸

PF-04447943 is a selective PDE9i that was developed for use in Alzheimer’s disease, but has also been investigated in SCD. Chronic administration of PF-04447943 reduces leukocyte–platelet aggregates and markers of endothelial activation in mice with SCD¹¹⁰ and a phase 1 clinical trial (NCT02114203) has been conducted to assess the safety, tolerability, pharmacokinetics, and pharmacodynamics of PF-04447943 in subjects with SCD. Improvements in the inflammatory markers, soluble E-selectin and heterocellular aggregates, were reported in patients on PF-04447943¹¹¹; however, although apparent safety was observed, some adverse events were also registered in patients in the PF-04447943 arms (vaso-occlusive crisis, biliary colic and pneumonia in 3 out of 22 patients). Another PDE9i, IMR-687, developed specifically for the treatment of SCD, displays a very high specificity for PDE9A, with a low brain penetration. Treatment of SCD mice with IMR-687 for 30 days inhibited microvascular stasis following hypoxia, decreased RBC sickling, elevated HbF-positive RBC, and reduced leukocytosis.¹¹² IMR-687 was granted Rare Pediatric Disease designation by the FDA and is currently being evaluated in a randomized, placebo-controlled, multicenter study phase 2 trial (NCT03401112) to determine the safety, pharmacokinetics, and preliminary pharmacodynamics of escalating doses of IMR-687 in patients with SCD.

Potential limitation of cGMP modulation therapy in SCD

One potential disadvantage of employing cGMP modulation therapy in SCD could be the role that NO and cGMP signaling may play in pain sensing. NO and cGMP participate in inflammatory and neuropathic pain processing. NO can act as a neurotransmitter, mediating peripheral and central nociception,¹⁵³ while neuronal or inducible NO synthase-derived NO and consequent cGMP-dependent neuronal signaling play a role in central sensitization, inducing pain hypersensitivity and contributing to inflammatory and neutropathic pain¹⁵⁴ Conversely, there is some evidence that NO may also have analgesic effects, mediating the analgesic effects of opioids and other substances.¹⁵³

Given that pain is a major complication of SCD, use of these therapies could be potentially problematic in this disease. Indeed, as previously mentioned, use of sildenafil in patients with SCD was unsuccessful due to an unexpectedly higher frequency of pain episodes in those patients in the sildenafil arm, compared with the placebo arm.¹⁰⁷ However, while myalgia and back pain have been associated with chronic PDE5 inhibitor administration,¹⁰⁷ such side effects have not yet been associated with guanylate cyclase activator/stimulator therapy for the treatment of heart failure,¹²⁴ although reports of musculoskeletal disorders for riociguat, compared with sildenafil and tadalafil, are higher during their use for pulmonary hypertension.¹⁵⁵ As such, pain evaluations will need to be carefully made when carrying out clinical trials of cGMP modulators in patients with SCD.

Conclusion

SCD is associated with a reduced life expectancy and lifetime morbidity. Hydroxyurea, the current mainstay of SCD pharmacotherapy, reduces mortality, transfusion requirement, VOE frequency and incidence of ACS via HbF elevation, and probably, via its immediate anti-inflammatory effects, all of which may be mediated by the ability of hydroxyurea to upregulate intracellular cGMP-dependent signaling. Despite the success of hydroxyurea therapy in SCD, investigations continue to identify compounds that can be used in combination with hydroxyurea, or alone, with a view to further reducing the frequency of painful VOEs and also for use following the onset of VOC and hospitalization of patients, as current therapeutic options for treating VOC are limited to pain medication and hydration. Pharmacological grade L-glutamine supplementation was recently approved by the FDA¹⁵⁶ and biological drugs such as crizanlizumab,¹⁵⁷ which demonstrated promising results in a recent clinical trial, show potential for use in SCD, due to their ability to reduce oxidative stress and abrogate P-selectin mediated cellular interactions, respectively. However, classes of drugs that upregulate cGMP-dependent signaling by providing NO under hypoxic conditions or by bypassing the need for NO delivery and amplifying intracellular cGMP concentrations may also constitute a major approach for use in SCD (see Table 1), in combination, or not, with hydroxyurea.

Authors’ contributions

NC and LT wrote the Review together. Both authors have read and approved the final manuscript.

DECLARATION OF CONFLICTING INTERESTS

NC has received research funding from Bayer AG and participates in technology registered with the INPI, Brazil, for use of a pharmaceutical composition of hydroxyurea and a PDE9 inhibitor for the control of vaso-occlusive processes.

FUNDING

The authors acknowledge grant support from FAPESP Brazil (Grant numbers 2018/08010–9, 2017/14594–0 and 2014/00984–3).