PerspectiveDrug Discovery

Gaseotransmitters: New Frontiers for Translational Science

See allHide authors and affiliations

Science Translational Medicine  24 Nov 2010:
Vol. 2, Issue 59, pp. 59ps54
DOI: 10.1126/scitranslmed.3000721


Translational research on endogenous gaseous mediators—nitric oxide, carbon monoxide, and hydrogen sulfide—has exploded over the past decade. Drugs that modulate either the gaseous mediators themselves or their related intracellular signaling pathways are already in use in the clinics, and still more are being tested in preclinical models and clinical trials. Discussed here are the chemical and pharmacological properties that present challenges for the translation of these potentially toxic molecules.

A 2008 article in Wired magazine (1) entitled “Hydrogen Sulfide May Kill Us, Bring Us Back to Life” helped to spark the interest of the lay public in the notion of using toxic gaseous substances in the clinic. But translational research on endogenous gaseous mediators has been expanding substantially over the past decade. A number of drugs already in clinical use and drug candidates now being tested in preclinical studies or clinical trials exert their effects either through the donation of the gaseous mediators nitric oxide (NO), carbon monoxide (CO), and hydrogen sulfide (H2S) or through modulation of their intracellular second-messenger pathways. Because of their crucial biological functions, gaseotransmitters offer diverse therapeutic opportunities. However, their unique chemical and pharmacological properties can also present unexpected challenges for translational science.

For many decades, NO was known as a toxic gas, familiar only to scientists interested in the physics of lightning and related atmospheric events or in the toxicology of car engine exhausts and smog. The discovery in the late 1980s that NO is produced by mammalian cells (2, 3) started a true revolution in the field of gaseotransmitters. NO is generated by a family of enzymes (NO synthases) from the amino acid l-arginine, and many fundamental regulatory functions of NO have been discovered in the cardiovascular, immune, and central and peripheral nervous systems (4, 5). A decade later, CO—another toxic gas known previously in toxicological and medical sciences as a dangerous byproduct of engines and furnaces—emerged as a neurotransmitter and cardiovascular and immune regulator. CO is produced as a result of heme metabolism via a family of enzymes called heme oxygenases (6, 7). Even more recently a third gas, H2S—previously recognized mainly as a toxic gas and environmental hazard produced in volcanic emissions, swamps, and certain industries—has rapidly emerged as an endogenous biological mediator with important regulatory roles in the nervous and cardiovascular systems and in the regulation of cell and whole-body metabolism (8, 9). H2S is produced from cysteine by the action of several enzymes: cystathionine-β-synthase (CBS), cystathionine-γ-lyase (CSE), and 3-mercaptopyruvate sulfurtransferase (3-MPST).

There are many similarities among these three gaseotransmitters in terms of their biology and pharmacology (Table 1 and Fig. 1). At physiological concentrations, all three gases act as vasodilators as well as broad-spectrum anti-inflammatory and cytoprotective agents. A number of therapeutic approaches at the clinical or preclinical stages of development are based on various aspects of gaseotransmitter pharmacology (10). These approaches, among others, include the therapeutic administration of (i) biological precursors or donors (11) of these gases, (ii) modulators of intracellular signaling pathways affected by these molecules, and (iii) the gaseotransmitters themselves (in inhaled gaseous form, as injected formulations, or as oral prodrugs) (Table 1).

Table 1. Therapy is a gas.

Therapeutic parameters of NO, CO, and H2S. For intracellular effector pathways, see Fig. 1. For additional information, see (8, 10, 42). ARDS, acute respiratory distress syndrome; COPD, chronic obstructive pulmonary disease.

View this table:
Fig. 1. Networking multitaskers.

Intracellular effector pathways of NO (blue), CO (red), and H2S (yellow) are shown. NO: NO donors or inhaled NO gas, as well as NO produced by NOS (nitric oxide synthase) from l-arginine activates sGC, which converts guanosine triphosphate (GTP) to cGMP. cGMP, in turn, acts on cGMP-dependent protein kinases (PKG) to mediate some of the biological responses to NO. Some of the other cellular effects of NO are related to activation (opening) of potassium-dependent adenosine triphosphate (ATP) channels (KATP channels) as well as other membrane channels (not shown). Other cellular effects of NO are mediated by S-nitrosylation and regulation of various redox pathways and responses. CO: Inhaled CO or CORMs supply CO to cells. The cells also produce CO from heme via the catalytic action of heme oxygenase. CO, in turn, activates various protein kinase–regulated signaling pathways (PK signaling); this effect occurs, in part, via inhibition of mitochondrial respiration and subsequent low-level induction of intracellular oxidants [reactive oxygen species (ROS)]. CO also opens KATP channels as well as other potassium and calcium channels (not shown) and is an activator of sGC (but is substantially less potent than NO in this respect). H2S: Inhaled H2S or H2S donors deliver biologically active H2S to cells. The cells also produce H2S from l-cysteine by the action of three distinct enzymes: cystathione-beta-synthase (CBS), cystathione-gamma-lyase (CSE), and 3-mercaptopyruvate sulfurtransferase (3-MPST). The cellular effects of H2S include inhibition of cytochrome c oxidase in the mitochondria, regulation of redox processes, and modulation of gene expression. H2S also opens KATP channels as well as other potassium and calcium channels (not shown). H2S inhibits phosphodiesterase (PDE) activity, and through this mechanism it can enhance the biological effects of NO-stimulated cGMP. This scheme does not represent all of the known effects or pathways of these three gaseotransmitters, and not all of these pathways are always present in a single cell type at the same time.


The first clinical introduction of a drug that acts as a donor of a gaseotransmitter preceded the rational understanding of the underlying biological processes by many years. Nitroglycerin and organic nitrates have been used as vasodilator and anti-angina compounds for over a century, although the discovery that they release NO, which then activates an endogenous second-messenger system [soluble guanylyl cyclase (sGC)/cyclic guanosine monophosphate (cGMP)] came about only in the 1970s (12). Furthermore, the realization that these drugs are, in effect, replacement therapies for the vasodilator endothelium-derived relaxing factor (EDRF) (also known as NO) came in the late 1980s (24).

An example of a modulator of intracellular signaling pathways affected by gaseotransmitters is the phosphodiesterase-5 (PDE5) inhibitor sildenafil, which enhances the biological actions of cGMP (the second messenger in the main signal transduction pathway of NO). Initially, sildenafil was studied in clinical trials aimed at cardiac protection. It was during these trials that the compound was serendipitously found to elicit an unexpected biological effect: induction of an erectile response (13). The delineation of the compound’s precise cellular mode of action only occurred subsequently. Sildenafil was shown to inhibit PDE5, which leads to the inhibition of cGMP degradation in the corpus cavernosum smooth muscle of the penis; this action is, in effect, an enhancement of the biological response to endogenously produced NO. A detailed understanding of this mechanism of action was essential for the rational design and development of next-generation PDE5 inhibitors (14).

A third clinical example is the use of inhaled NO gas as a therapy for pulmonary hypertension—a pressure increase in pulmonary blood vessels. In this case, the introduction of NO gas into clinical use was the result of a rational mechanism-based drug testing and development approach. Fundamental laboratory research observations made in the mid-1990s suggested that there are insufficient amounts of biologically active NO in the pulmonary vasculature in animals with severe pulmonary hypertension. Thus, it was hypothesized that therapeutic replacement of NO, via the inhalation route, may provide substantial benefit in newborn infants with primary pulmonary hypertension. This hypothesis was subsequently confirmed, first in preclinical studies and later in a series of clinical studies (15).

Many aspects of the pharmacology and experimental therapy of gaseotransmitters are radically different from the characteristics of small-molecule therapeutics. These aspects provide interesting opportunities and challenges for basic research and translational science and are discussed in the subsequent sections.


The three gaseotransmitters NO, CO, and H2S represent some of the smallest biologically active molecules in existence. They diffuse easily through cell membranes, reaching their multiple intracellular targets. Inhalation of these gases can produce significant local (intrapulmonary) actions and, in some cases, systemic effects as well. In the case of inhaled NO, the original hypothesis was that the majority of the biological effects would occur inside the lung tissue (producing selective pulmonary vasodilatation), because the reaction of NO with hemoglobin after reaching the blood was thought to inactivate the molecule (15). However, subsequent work demonstrated that the oxidation products of NO (nitrite and nitrate) as well as some of its other reaction products (for example, S-nitrosothiols) can exert remote biological effects. Thus, inhaled NO has recently been shown to exert beneficial effects in exploratory trials in patients with liver transplantation or limb ischemia (16, 17). In the case of CO, the inhaled gas binds to hemoglobin in the red blood cells, which may serve as a CO carrier, delivering the molecule to tissues (18, 19). For inhaled H2S, a recent study compared the amounts of biologically active, reactive sulfide species in the blood in response to inhaled H2S versus infusion of an H2S-donating formulation, thereby establishing a de facto bioequivalency between the two modes of delivery (20).


In theory, endogenous molecules may have a distinct advantage when applied as therapeutics, because the body “already knows” how to deal with them; the cellular responses are predictable, and specific elimination pathways are present. In some ways, gaseotransmitter therapy may be viewed as hormone-replacement therapy of an exceptional kind. Nevertheless, as with all therapeutic agents (including hormone therapy, and especially with gases that can be viewed as both toxic agents and biological regulators), it is important to determine at which point the elimination pathways become overwhelmed and when the therapeutic response turns into a toxicological effect.

Because all three gaseotransmitters are produced endogenously as part of the human body’s normal physiology, there are significant basal concentrations of these gases in many cells and tissues. This scenario is quite different from that encountered with most synthetic small-molecule drugs. An added complexity is that endogenous amounts of all three gaseotransmitters may increase or decrease in various disease conditions. On one hand, these changes may be interesting for diagnostic purposes. On the other hand, the possibility exists that such changes may affect the overall biological response to the exogenously administered gaseotransmitter. For instance, it is well known that the vasodilator response to NO donors is potentiated in blood vessels after pharmacological inhibition of endogenously produced NO (21).


Therapeutic discovery with gaseotransmitters brings to mind the alleged advice of baseball legend Yogi Berra: “When you come to a fork in the road, take it.” Typically, one of the key requirements for small-molecule therapeutic agents is to have a clearly defined, specific mode of action (for example, blockade of a single receptor subtype or selective inhibition of a particular enzyme isoform). In contrast, gaseotransmitters are natural multitaskers: They have evolved to act on multiple cellular targets, in order to induce complex and coordinated biological responses. The cellular targets and pathways affected by the gaseotransmitters (in many cases, in a concentration-dependent and cell- or tissue-dependent fashion) can create challenges for translating laboratory observations into therapeutic applications.

Key checkpoints in the biological effects of gaseotransmitters are amenable to therapeutic intervention. For NO, the principal pathway of smooth muscle relaxation (and hence, increase in blood flow, lowering of blood pressure, or penile erection, depending on the type of smooth muscle involved) requires the activation of the soluble guanylyl cyclase enzyme by the interaction of NO with an iron molecule on the enzyme’s heme prosthetic group. This binding is followed by the elevation of intracellular concentrations of cGMP, which results in the activation of cGMP-dependent protein kinases or the S-nitrosylation of proteins (24). In addition to measuring the physiological responses, various parameters (biological markers) related to these pathways can be monitored (for instance, circulating cGMP concentrations) (22).

The cytoprotective/anti-inflammatory effects of NO (as well as of CO and H2S) involve multiple complex pathways, a fact that underlines the importance of appropriate biomarker identification. At high local concentrations, the cytotoxic effects of NO, CO, and H2S are orchestrated via several interrelated pathways, including a direct inhibition of mitochondrial respiration (Fig. 1) (49). In the case of NO, an additional pathway of cytotoxicity involves the formation of peroxynitrite, a highly reactive species formed by the rapid reaction of NO with superoxide (23).

The therapeutic effect/toxic side effect concept may become particularly challenging when the therapeutic mode of action (for example, interaction with cytoprotective or anti-inflammatory kinase pathways) and the toxicity (such as the inhibition of mitochondrial respiration) stem from the very same molecular target. For instance, in the case of CO, it has been demonstrated that the inhibition of cytochrome c oxidase (followed by the inhibition of mitochondrial respiration and generation of reactive oxygen species) is responsible for both the toxic effects of this gas and for the therapeutic modulation of anti-inflammatory signal transduction (24). In such cases, the direction of the net biological effect (protection or toxicity) may be a function of the degree of cytochrome c inhibition.


A characteristic aspect of gaseotransmitter pharmacology is the bell-shaped dose-response relationship curve. The beneficial (for example, cytoprotective or anti-inflammatory) effects of NO, CO, or H2S typically occur at low (near-physiological) concentrations, whereas at higher concentrations the effects diminish and toxicity may ensue. For instance, murine studies have demonstrated that the reduction of myocardial infarct size by nitrite (a precursor of NO) or by H2S occurs at certain well-defined doses; upon further elevation of the dose, the protection is lost, or the compound may even worsen the outcome (25, 26). Of course, we have known at least since the time of Paracelsus (1493–1541) that “all drugs are poisons; the benefit depends on the dosage.” But the exact definition of the therapeutic dose range and careful monitoring for signs of potential toxicity at higher exposures is especially relevant for gaseotransmitters. Although the toxicology of all three gases has been studied extensively for many decades in the context of environmental and medical sciences, the very chemical nature of NO, CO, or H2S may lead to reservations against the general idea of using such well-known toxic gases (or compounds that contain or release them) for therapeutic purposes.


The gaseous nature of NO, CO, and H2S allows a convenient mode of delivery via inhalation. In fact, the only U.S. Food and Drug Administration (FDA)–approved use of an inhaled therapeutic gaseous drug is inhaled NO, which is the first line of therapy for primary pulmonary hypertension of the newborn (15). The lung, however, represents a “two-way street” for gaseotransmitters. In all three cases (NO, CO, and H2S), healthy humans exhale substantial amounts of these gases (6, 7, 2729); this physiological process can be used either as a way to monitor exposure to infusion of prodrugs of the gaseous transmitters, as recently shown in the case of intravenously administered H2S (29), or as a diagnostic marker (26). For instance, an increase in the amounts of exhaled NO has been used in the diagnosis of asthma; exhaled NO concentrations correlate well with the degree of eosinophilic inflammation and have also been used to monitor the therapeutic effectiveness of glucocorticoid therapy (30, 31). The measurement of exhaled CO concentrations is being explored for the diagnosis of asthma, chronic obstructive pulmonary disease, lung transplantation, and hemoglobinopathies (3235).

During gaseotransmitter therapy, the therapeutic agent is the gas itself. Thus, the delivery of the gaseous molecule (or a prodrug that releases it) may occur via inhalation or by injection or oral administration, whereas the actual biological effect is mediated by a soluble gas, which diffuses into the various compartments of the body. These characteristics of gaseotransmitters require a complex integration of pharmacology and biology when considering dose-responses (which are often bell-shaped), biological exposure (which is determined by delivery and excretion), therapeutic indices (which are often narrower than those of small-molecule drugs), and metabolites (which can be biologically active and present in untreated people because of the endogenous baseline production of gaseotransmitters).

In some cases, targeting downstream effector pathways of gaseotransmitters with classical small-molecule drug candidates can reduce some of the above-mentioned complexities and may represent a viable alternative approach to the direct administration of the gaseotransmitters or their prodrugs. For instance, inhibitors of PDE5 enhance cGMP-mediated responses (13, 14). The fact that a specific PDE isoform is present in the corpus cavernosum tissue provides the opportunity for tissue selectivity, thus avoiding overt systemic effects such as the lowering of blood pressure.

An emerging class of clinical-stage compounds that stimulate soluble guanylyl cyclase (3638) are being developed that either bind to the NO-responsive heme group of the soluble guanylyl cyclase enzyme or, in case of the oxidized form of the enzyme, which is missing the heme group, that mimic the conformation of the absent heme group on the enzyme, thereby reactivating it. Selected members of these compound classes are now being tested in clinical trials, such as the sGC stimulators riociguat for pulmonary hypertension and cinaciguat for acute decompensated heart failure (39, 40).


Prodrugs and donors of NO have been used for many decades for the treatment of angina pectoris and hypertension, as well as for many other systemic and topical indications (5, 8). For CO, several series of boron- or ruthenium-based compounds (CO-releasing molecules, or CORMs) have been synthesized in the past decade in order to provide an injectable or oral mode of therapeutic CO administration. CORMs exert potent therapeutic effects in preclinical studies of (i) ischemia-reperfusion injury of various organs (such as the heart and kidney), in the contexts of both thrombosis and organ transplantation; (ii) several forms of acute and chronic inflammation (including arthritis, colitis, and neuroinflammation); and (iii) a variety of other diseases such as postoperative ileus and sepsis (8, 41, 42).

A distinct approach for the therapeutic exploitation of NO and H2S relates to the synthesis of modified small molecules with NO- or H2S-releasing functional groups (10). This approach recently culminated in phase III studies with the nonsteroidal anti-inflammatory drug Naproxcinod (nitronaproxen). The rationale for creating this compound was that incorporation of an NO-donor group into the structure of naproxen may (i) produce a local, NO-mediated vasodilatation that might protect the stomach from the ulcerogenic effect of the parent compound, and (ii) exert a slight, potentially therapeutic vasorelaxant/hypotensive effect (43). Although the U.S. FDA recently declined the new drug application for Naproxcinod submitted by the French company NixOx (44), it is possible that in the future, similar conceptual approaches may find their way into clinical use. Similar to the combined NO donors, H2S-releasing nonsteroidal anti-inflammatory drugs as well as H2S-releasing PDE inhibitors have been synthesized. Some of these compounds show distinct pharmacological profiles, such as therapeutic advantages (higher efficacy or improved side-effect profiles) over the parent compound (10, 45).


There are many examples in which hitherto unknown gaseotransmitter-related modes of action have been identified for various well-established drugs. For instance, part of the therapeutic effect of cholesterol-lowering statin drugs is now attributed to the activation and enhanced expression of the endothelial isoform of NO synthase, thereby improving cardiovascular function (46). The PDE5 inhibitor tadalafil is another interesting example; some of its biological effects in the heart have recently been shown to occur via the stimulation of H2S production in vivo (47).

The rapid progress of basic and applied science in the field of gaseotransmitters is both exciting (for investigators working in the field) and challenging (for those who attempt to integrate the massive amount of rapidly changing information into clinical practice). As a result of intensive efforts from thousands of scientific groups worldwide, many notions about gaseotransmitters that were thought to be well established need substantial revision. For instance, nitrite (considered an inactive metabolite of NO) has emerged as a biologically active prodrug for NO, opening up a wide range of possible therapeutic applications for the treatment of conditions such as ischemic heart disease, peripheral artery disease, hypertension, transplantation, and others (25, 48, 49). In addition, researchers recently discovered potent antibacterial effects of CORMs, with obvious therapeutic implications (50). Another recent study demonstrated that topical administration of H2S promotes angiogenesis, which may have therapeutic uses in wound healing (51); whereas other new work identified H2S as an “endogenous sildenafil”—that is, an inhibitor of PDE and an enhancer of the effects of endogenously produced NO (52)—a finding that will necessitate a reevaluation of the biological functions of both NO and H2S.

The science of gaseotransmitters has thus been rapidly progressing in recent years (Fig. 2) and remains an exciting area for translational research. It is safe to predict that continuing research in this scientific area will lead to the identification of new biological concepts with therapeutic and diagnostic potential.

Fig. 2. Milestones in gaseotransmitter translation.




  • Citation: C. Szabo, Gaseotransmitters: New Frontiers for Translational Science. Sci. Transl. Med. 2, 59ps54 (2010).

References and Notes

  1. Funding: The academic research of the author in the field of gaseotransmitters is supported by NIH grant R01GM060915 and a grant from the Shriners Burns Hospitals. Competing interests: The author is a former officer and current shareholder of Ikaria Holdings, a clinical-stage biopharmaceutical company involved in research, development, and clinical therapy in the area of gaseotransmitters.

Stay Connected to Science Translational Medicine

Navigate This Article