© 2003 Annual ReviewsAll images redrawn from D. Boehning, S.H. Snyder, Ann Rev Neurosci, 26:105–31, 2003.

Neuronal nitric oxide synthase (nNOS) is localized to NMDA receptors by the PDZ-domain adaptor protein PSD95. Calcium entry activates nNOS by a calcium/calmodulin-dependent mechanism. NO can diffuse to neighboring cells to activate soluble guanylyl cyclase or to nitrosylate cysteine residues on target proteins. Nitrosylation inhibits NMDA receptors providing a negative feedback loop. CAPON competes with PSD95 for nNOS to facilitate the nitrosylation and subsequent activation of Dexras1.

High levels of carbon monoxide interfere with cellular respiration and pollute the environment. Hydrogen sulfide (H2S), another chemical asphyxiant, paralyzes the sense of smell and at lower levels produces the rotten-egg stink prized by children using their first chemistry sets. But even the noxious to the downright deadly can have a subtler side: at minute concentrations, both gases transmit biological signals between cells....


Gasotransmitters are conserved throughout evolution and produce numerous effects. Insects, marine sponges, plants, and bacteria all use NO as a signaling molecule. In plants, NO modulates leaf expansion and root growth, and it also seems to protect plants from environmental and infection stresses.4 Moreover, says Marie-Alda Gilles-Gonzalez, at the University of Texas Southwestern Medical Center at Dallas, cells often use gases to signal their positions. Differential expression of the gas transmitter can tell the cells in an embryo which way is up, for example. "Positioning is clearly important in embryonic development, tumor formation, and symbiotic as well as pathogenic relationships," she comments.

NO exemplifies the functional diversity of gasotransmitters. NO dilates the vasculature, controlling regional blood flow and pressure. (Excessive production can produce hypotension in septic shock.) Macrophages, neutrophils, and leukocytes use NO to fight viruses, bacteria, and parasites. In the central nervous system, NO contributes to memory, pain, and appetite.

Carbon monoxide is equally pleiotropic. Living cells produce CO as a by-product of heme degradation, a process catalyzed by heme oxygenase (HO) enzymes. "Long thought to represent metabolic waste, endogenous CO attracted much recent attention as a potential physiological regulator," says Stefan Ryter at the University of Pittsburgh Medical Center. Indeed, Roberto Motterlini, at Northwick Park Institute for Medical Research in London, adds that the HO pathway and two products, CO and biliverdin, are important components of tissue response to stressful stimuli and essential for restoring cellular homeostasis.

Heme oxygenase-1 (HO1) seems to protect cells against oxidative stress and apoptosis. HO1 also reduces several hallmarks of inflammation including edema, leukocyte migration, and production of pro-inflammatory cytokines. NO seems to interact with cyclooxygenase (COX) pathways,5 which are responsible for producing pro-inflammatory prostaglandins. For this reason researchers at the University of Pittsburgh are investigating whether measuring gases in exhaled breath offers a diagnostic marker of inflammation.

Although the evidence supporting H2S as a gasotransmitter is not as extensive as that for CO and NO, it continues the theme of functional diversity. Vascular smooth muscle cells, for instance, generate H2S, catalyzed by a specific enzyme, from L-cysteine.6 Hideo Kimura, of the National Institute of Neuroscience in Japan, comments that H2S modulates synaptic activity in the brain, regulates the activity of glia and, according to research currently in press, protects neurons from oxidative stress. Kimura adds that H2S also enhances responses mediated by NMDA receptors (which bind excitatory amino acid transmitters), including the induction of long-term potentiation, a key step in memory formation.


Gasotransmitters contribute to such diverse functions because they can diffuse easily into cells and modulate the function of key proteins. Solomon Snyder at Johns Hopkins Medical School explains that NO and CO diffuse into adjacent cells, where they activate guanylyl cyclase.2 This enzyme catalyzes the formation of the second messenger, cyclic GMP. More recent evidence suggests that NO performs a post translational S-nitrosylation on cysteine residues of proteins as diverse as glutamate receptors, alpha-tubulin (the main component of microtubules), the sodium pump, and metabolic enzymes including glyceraldehyde-3-phosphate dehydrogenase. "Nitrosylation is an extraordinary and remarkably effective way of conveying information," Snyder says. Nitric oxide synthase (NOS) binds to carrier proteins, which, in turn, bind targets for nitrosylation. This cascade means that nitrosylation is selective.


© 2003 Annual ReviewsAll images redrawn from D. Boehning, S.H. Snyder, Ann Rev Neurosci, 26:105–31, 2003.

Increases in intraneuronal Ca2+ through voltage gated Ca2+ channels (VGCC) or release from intracellular stores activate PKC. Through phosphorylation, PKC activates CK2 which activates HO2 on the endoplasmic-reticulum (ER) membranes. HO2 cleaves heme to generate CO, iron, and biliverdin in a reaction dependent on cytochrome P450 reductase (CPR). CO can then activate soluble guanylyl cyclase (SGC) and Ca2+-activated K+ channels (KCa)

Jaggar notes that CO only weakly activates soluble guanylyl cyclase. One possibility under investigation is that so-called permissive factors enhance the activation of guanylyl cyclase by CO. The gas may also directly activate large-conductance calcium-activated potassium channels and some researchers believe that NO may be necessary for CO to elicit its cellular effects. "These questions will be investigated over the next few years," Jaggar says

To better understand gaso-transmitter mode of action, Ryter is trying to identify novel targets, such as those containing heme or iron, that may provide cellular binding sites for CO, the so-called gas sensors.7 "We are interested in exploring the signaling consequences of such targets with respect to modulation of enzymatic activities or intracellular redox potential," he says.

Meanwhile, Gilles-Gonzalez's group contributed to the recognition that gas sensors such as heme, guanylyl cyclase, and the heme-containing transcription factor CooA are sufficiently precise to distinguish diatomic gases. Their research revealed that most gasotransmitter sensors share a modular design. First, the proteins have a sensory module that contains a heme cofactor. "This region is reminiscent of myoglobin, and it is where the gas binds," Gilles-Gonzalez says. "Most proteins lack the functionality to bind O2, CO, and NO with their amino acid side chains, and so they co-opt a heme cofactor."

Second, sensory proteins have a transmitter module that amplifies the gas-binding signal and converts it to a language familiar to the cell, Gilles-Gonzalez says. For example, some sensors are kinases that phosphorylate other proteins. Others are cyclases that convert a nucleotide to a cyclic form. Still others are phosphodiesterases that bind transcription factors or signal the chemotaxis machinery so that cells move toward or away from a gas.

Gilles-Gonzalez's lab studies several O2 and CO sensors from diverse species, including an O2-sensing phosphodiesterase that controls biofilm formation by the bacterium Acetobacter xylinum and a CO-sensing transcription factor that controls circadian rhythm in mammalian brains. "Remarkably, the sensory modules in all of these proteins have a similar architecture, even though their transmitter modules differ radically from each other," she says. "We especially want to understand how the sensory and transmitter modules in sensors communicate with each other. We are working on solving the structures of these molecules and elucidating the dynamics of their interactions."


The discovery of gasotransmitters challenged some fundamental assumptions about cell signaling. "These new messenger molecules challenge old dogma about just what is a neurotransmitter," Snyder says. He notes that nerves fire in response to rapid and unpredictable impulses. So, researchers assumed that transmission required large pools of transmitter with only 1%-2% released from the synaptic vesicles with each impulse.

<p>H<sub>2</sub>S SYNTHESIS:</p>

© 2003 Annual ReviewsAll images redrawn from D. Boehning, S.H. Snyder, Ann Rev Neurosci, 26:105–31, 2003.

Calcium entry through glutamate receptors activates the H2S-synthesizing enzyme, cystathionine β-synthase (CBS), via calcium/calmodulin. H2S activates K+ channels and/or adenylate cyclase.

"Yet, NO and CO cannot be stored in vesicles," Snyder remarks. "Hence, their biosynthetic enzymes must be exquisitely regulated to permit activation with each nerve impulse." Calmodulin helps cells attain this regulation. NO synthase and HO bind calmodulin, and neuronal depolarization triggers calcium entry into cells. The calcium binds to and activates both calmodulin and the enzyme.

Furthermore, Ryter adds that carbon monoxide is relatively stable, so it does not conform to the classical definition of a transmitter as degrading rapidly. "The mechanisms by which the CO-generated signal is attenuated, and by which CO is cleared from the site of action remain unanswered," he says.


Given the diverse range of actions, it is perhaps not surprising that research into gaso-transmitters has opened rich therapeutic opportunities. Indeed, some drugs that modulate gasotransmitters are already used clinically. NO donor compounds, such as sodium nitroprusside and glyceryl trinitrate, have been used clinically as vasodilators. Sildenafil, tadalafil, and vardenafil, the drugs used to treat erectile dysfunction, act by inhibiting phosphodiesterase 5. This enhances NO's ability to increase cGMP concentrations and, therefore, dilates the penile vasculature.

Darren Boehning at Johns Hopkins speculates that, in the future, disorders characterized by dysfunctional gastrointestinal motility might represent a potential market for drugs modulating gasotransmitter function. He is a member of a team investigating the cotransmitter roles played by NO and CO in the gut; they also are unraveling HO2's association with other brain proteins. "There may also be a market for psychoactive drugs, since NO has been linked to aggressive behavior," he says. Furthermore, CO is required for ejaculation and anal sphincter function, while HO2 activity seems to be neuroprotective. Such findings open yet more therapeutic opportunities.

Delivering gas by inhalation offers one novel approach to exploiting these opportunities. Furthermore, prototype drugs modulating gasotransmitters other than NO are beginning to emerge. In the United Kingdom, for example, Motterlini's team developed a novel means to deliver therapeutic CO: a group of transition-metal carbonyls.8 "Our attempt to diversify the portfolio of CO-releasing molecules that possess a variety of chemical characteristics, such as water-soluble versus lipid-soluble and slow versus fast releasers, could facilitate the development of drugs for the therapeutic delivery of CO in a safe, measurable, and controllable fashion."


Inevitably, numerous questions remain. Boehning asks for example, whether physiologic CO levels activate soluble guanylyl cyclase or act through other targets, such as potassium channel activation. He also wants to know whether there is adequate heme for HO to generate sufficient CO on demand during repeated neuronal depolarizations. "Another interesting question is how NO and CO interplay in tissues in which both appear to be neurotransmitters," he adds. "Why do you need both?"

Motterlini agrees that understanding why the body needs to produce more than one gaseous molecule and characterizing interactions between the gasotransmitters are key issues for future research. "Do they act in a concerted way, or is one pathway controlling the other? And if so, why?" he asks. "NO is more potent than CO, thus why is CO needed?" Yet, he argues, the two gasotransmitters may show different specificities.

Gasotransmitters will undoubtedly be an area of intense research over the next few years. And the number of gases shown to act as endogenously produced transmitters or showing pharmacological actions at low doses is likely to grow. Recent evidence suggests that ammonia is a vasoconstrictor, possibly by acting through intracellular alkalinization. Sulfur dioxide and nitrous oxide (N2O), both produced by bacterial metabolism, may be a vasodepressor and NMDA antagonist, respectively.6 "Gases commonly known for their noxious effects at relatively high concentrations are produced by the body continuously and in minute quantities and are still capable of exerting crucial physiological activities," Motterlini concludes. "In reality, rather than an overturned dogma, it is a validation of a principle propounded by Paracelsus in the 16th century." The Swiss physician and father of pharmacology said that all substances are poisons; it's only the dose that differentiates a poison from a remedy.

Mark Greener mgreener@the-scientist.com

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?