Isoprene decreases the concentration of nitric oxide in leaves exposed to elevated ozone


Author for correspondence: Francesco Loreto Tel: +39 06 90672530 Fax: +39 06 9064492 Email:


  • • Isoprene reduces visible damage (necrosis) of leaves caused by exposure to ozone but the mechanism is not known. Here we show that in Phragmites leaves isoprene emission was stimulated after a 3-h exposure to high ozone levels.
  • • The photosynthetic apparatus of leaves in which isoprene emission was inhibited by fosmidomycin became more susceptible to damage by ozone than in isoprene-emitting leaves. Three days after ozone fumigation, the necrotic leaf area was significantly higher in isoprene-inhibited leaves than in isoprene-emitting leaves.
  • • Isoprene-inhibited leaves also accumulated high amounts of nitric oxide (NO), as detected by epifluorescence light microscopy.
  • • Our results confirm that oxidative stresses activate biosynthesis and emission of chloroplastic isoprenoid, bringing further evidence in support of an antioxidant role for these compounds. It is suggested that, in nature, the simultaneous quenching of NO and reactive oxygen species by isoprene may be a very effective mechanism to control dangerous compounds formed under abiotic stress conditions, while simultaneously attenuating the induction of the hypersensitive response leading to cellular damage and death.


Isoprene is one of the most abundant volatile organic compounds emitted by plants. Isoprene drains a considerable percentage of the carbon fixed through photosynthesis out of the pathway forming structural and storage sugars, especially in stressed leaves (Sharkey & Yeh, 2001). There has been considerable interest in investigating the possible eco-physiological role of isoprene emission. Two general types of hypotheses have been advanced: those hypothesizing a regulatory role for isoprene in plant biochemistry, dissipating excess reductant, energy or carbon (Logan et al., 2000); and those hypothesizing a protective role of isoprene against environmental stresses. Preliminary observations showed that isoprene emission can be stimulated after recovering from environmental stress (Sharkey & Loreto, 1993). More recently, numerous studies have shown that isoprene can make leaves more tolerant to heat (Sharkey & Singsaas, 1995; Singsaas et al., 1997; Sharkey et al., 2001) and ozone (Loreto et al., 2001), perhaps by a similar mechanism. Isoprene can stabilize membranes defending the lipid bilayer by denaturation consequent to heat (Sharkey, 1996) or oxidative pressure (Loreto & Velikova, 2001), or may scavenge reactive oxygen species (ROS) in the gaseous phase, therefore lowering their pressure over membranes under stress conditions (Loreto et al., 2001).

We showed that isoprene fumigation reduced necrosis of ozonated leaves (Loreto et al., 2001). A unique signalling system is at the basis of hypersensitive response and cellular suicide under a wide range of biotic and abiotic stress (Beligni & Lamattina, 2001), including ozone exposure (Rao & Davis, 2001; Pasqualini et al., 2003). This involves release of nitric oxide (NO), and interaction of this volatile molecule with increasing levels of ROS, particularly H2O2 (Delledonne et al., 2001). Remarkably, no cellular death occurs when either one of these precursors (NO and H2O2) is not produced at high rates (Delledonne et al., 2001). In fact, a moderate increase of NO may per se promote activation of genes and formation of metabolites involved in plant defense (Delledonne et al., 2003), efficiently scavenging products of oxygen photoreduction (Beligni & Lamattina, 2002), and reducing oxidative damages to lipids, proteins and RNA (Wink et al., 1993; Beligni & Lamattina, 2002).

Isoprene reduces the amount of H2O2 formed in ozonated leaves (Loreto & Velikova, 2001), and quenches singlet oxygen (Affek & Yakir, 2002), thus offering a general protection against ROS. Isoprene could also react with NO or with peroxynitrite (ONOO) formed by NO–ROS interaction, reducing the presence of these compounds which may be harmful for many biological molecules (Lipton et al., 1993), and also indirectly modulating the amount of NO reacting with ROS to initiate cellular hypersensitive responses. We designed an experiment to test this hypothesis. We show that endogenous isoprene in fact quenches the amount of NO of Phragmites leaves, thus preventing the condition that triggers cell death, and preserving leaf physiological functions under ozone exposure.

Materials and methods

Plant material, protocols and statistics

Reed (Phragmites australis L.) plants were grown in 10-l pots filled with loamy soil in a controlled-environment cabinet (Fitotron, Sanyo-Gallenkamp, Leicester, UK). Pots were irrigated daily to soil water capacity to maintain the aquatic plants under unstressed conditions. For the same reason, the relative humidity of the air was maintained at > 80%. The air temperature was maintained at 30°C during the 14-h photoperiod and at 25°C during the dark period. The light intensity at the level of the leaves was c. 700 µmol m−2 s−1.

Measurements on single leaves were carried out after exposure to ozone (ozonated) with or without previous treatments with the following chemicals, alone or in different combinations: the isoprene-inhibitor fosmidomycin, the NO-donor sodium nitroprusside (SNP), and the NO-scavenger 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (CPTIO) (see details about treatments on the following subheadings). Control leaves were maintained in the gas-exchange cuvette for 4 h, in the same environmental conditions of all other leaves, and were fed with none, with either one of the chemicals or with both fosmidomycin and either SNP or CPTIO, but were not ozonated. Leaf extracts were used to assess in vitro NO generation by SNP and NO–singlet oxygen interaction. All treatments were replicated on at least six different leaves. Means ±se are presented unless otherwise stated. Means were statistically separated by Tukey's test, at probability for no difference P < 5% or 10%.

Gas-exchange measurements

Single leaves were cut from plants and their basal part was rapidly immersed in water to avoid water stress. The medium and apical part of the leaves were inserted in a 5 cm2 gas-exchange cuvette entirely coated with a Teflon film to avoid O3 uptake by cuvette materials, and exposed to a flow rate of 300 µmol s−1 of contaminant-free (O3, NO and isoprene-free) synthetic air made by mixing O2, N2 and CO2 (20%, 80% and 370 µg g−1, respectively). The leaves were maintained at a relative humidity of 80% and a leaf temperature of 30°C, as explained elsewhere (Loreto & Velikova, 2001), and were illuminated with a OSRAM Power Star lamp (OSRAM, Milano, Italy) generating an actinic light intensity of 700 µmol photons m−2 s−1 incident on the leaves. The air entering and leaving the cuvette was analyzed for its content of CO2 and H2O with a Li-Cor 6262 infrared gas analyzer (Li-Cor, Lincoln, NE, USA), and photosynthesis and stomatal conductance to water were calculated with the formulations of von Caemmerer and Farquhar (1981).

A part of the air leaving the cuvette was diverted into a gas-chromatograph (Syntech Spectras BTX Analyser GC 855, Syntech Spectras, Groningen, the Netherlands) for online analysis of the isoprene released by the leaf. The system details are described in Loreto et al. (2001) and Loreto and Velikova (2001).

Ozone treatment

When photosynthesis, stomatal conductance and isoprene emission were stable, c. 1 h after inserting the leaves in the cuvette, the leaf disc was fumigated with 300 ng g−1 of O3 for 3 h. Ozone was generated by passing a part of the O2 entering the cuvette through an UV light source (Ozonomatic, Rome, Italy). The rest of the synthetic air was then mixed immediately before reaching the gas-exchange cuvette. This way the formation of nitrogen oxidative products by direct reaction with active oxygen did not occur. The O3 concentration inside the cuvette was continuously monitored with an UV Photometric O3 analyzer (1008 Dasibi Environmental Corp., Glendale, CA, USA). Ozone visual damage was assessed by measuring with a scanner the necrotic (brown) and nonnecrotic areas in leaves 3 d after the treatment. These areas were expressed as a percentage of the total leaf area exposed to the treatment.

Isoprene inhibition by fosmidomycin

In some leaves, isoprene emission was inhibited by fosmidomycin, a compound that specifically blocks the pathway of isoprenoid synthesis in chloroplasts (Zeidler et al., 1998), before starting the ozone treatment. Fosmidomycin was fed to the leaves as an aqueous solution (5 µm) which was taken up through the transpiration stream. Isoprene emission was inhibited to a constant level (c. 10% of the original level) within 40–60 min after fosmidomycin feeding. In our previous experiments we tested if fosmidomycin per se has any effect on photosynthesis. No changes in photosynthesis, photosynthetic electron transport rate, or content of nonvolatile isoprenoids were observed for 3 h after fosmidomycin feeding in isoprene (Loreto & Velikova, 2001) and monoterpene-emitting leaves (Loreto et al., 2004).

NO localization in the leaves

NO was localized in the mesophyll cells of control (nonozonated) leaves and in leaves exposed to ozone, with or without previously inhibiting isoprene emission by feeding fosmidomycin. All samples were collected immediately after the treatment. NO (yellow–green) fluorescence was detected by epifluorescence light microscopy (excitation 495 nm; emission 515 nm; long-pass filter of 515 nm; Leica DM RHC microscope, Leica, Wetzlar, Germany) in leaf sections incubated in the dark with 10 µm 4,5-diaminofluorescein diacetate (DAF-2DA) for 3 h at 25°C (Kojima et al., 1998). DAF-2DA permeates through the cell membrane and is hydrolyzed to yield DAF-2, which is retained in the cell owing to its relatively poor permeability. Loaded DAF-2 reacts with NO to form fluorescent DAF-2 T. This method is sensitive for NO concentrations above 5 nm and the fluorescence intensity is linearly related to NO concentrations above this threshold.

NO formation by sodium nitroprusside and NO scavenging by 2-(4-carboxyphenyl)- 4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide

SNP (sodium nitroferricyanide dihydrate, NaFe(CN)5NO·2H2O) belongs to the group of organic nitrates, which are the most widely used NO donors. CPTIO scavenges NO, forming NO2 radical. Several concentrations of SNP and CPTIO were tested and we determined that 3 µm SNP and 7 µm CPTIO were the maximal doses to be used without harmful effect on photosynthesis during the time of the treatment. SNP or CPTIO were fed on some leaves as an aqueous solution through the transpiration stream, as for fosmidomycin. Ozone fumigation was started 60 min after SNP or CPTIO feeding. In the experiments with isoprene-inhibited leaves, firstly isoprene emission was inhibited by fosmidomycin and then SNP or CPTIO were fed.


Exposure to ozone caused a significant increase of isoprene emission, monitored immediately after the ozone treatment (Fig. 1). This effect was also observed in ozonated leaves in which NO content was enhanced by previously feeding the NO-donor SNP, or quenched by feeding the NO-scavenger CPTIO (Fig. 1). In leaves in which isoprene was inhibited by fosmidomycin before the ozone treatment, with or without concurrent application of SNP or CPTIO, the residual emission of isoprene was also slightly stimulated by ozone, but this increase was not statistically significant (Fig. 1).

Figure 1.

Emission of isoprene in nonozonated (controls, a) and ozonated (b) Phragmites leaves. Emission was detected immediately after the 3-h exposure to air containing 300 ng g−1 of ozone or, in nonozonated leaves, after maintaining the leaves for 3 h in the cuvette under ozone-free air. Emission is also reported for leaves previously fed with fosmidomycin (F) to inhibit isoprene formation, with sodium nitroprusside (SNP) to generate nitric oxide (NO), with 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (CPTIO) to scavenge NO, and to fosmidomycin + NO donor or NO scavenger. Means ±se (n = 6) are shown. Means were separated by Tukey's test and different letters identify means statistically different with P < 5% (double letters) or P < 10% (single letter).

Ozone reduced photosynthesis (Fig. 2a,b) and stomatal conductance (Fig. 2c,d), but the effect was exacerbated in leaves in which isoprene was previously inhibited by fosmidomycin. These leaves also showed a larger percentage of necrotic leaf area 3 d after the treatment, in comparison to ozonated leaves emitting isoprene (inset to Fig. 2b).

Figure 2.

Photosynthesis (a, b) and stomatal conductance (c, d) of controls (nonozonated, left-hand panels) and ozonated (right-hand panels) Phragmites leaves (open bars). Emission is also reported for controls and ozonated leaves previously fed with fosmidomycin to inhibit isoprene formation (closed bars). The inset to (b) shows the percentage of damaged (necrotic) leaf area with respect to the leaf area exposed to ozone in leaves emitting isoprene (open circle) and in isoprene-inhibited leaves (closed circle). Means ±se (n = 6) are shown. Means were separated by Tukey's test and different letters identify means statistically different with P < 5% (double letters) or P < 10% (single letter).

We used a NO-sensitive fluorescent probe (DAF-2 DA) to establish whether P. australis mesophyll cells actually produced NO and if NO was efficiently scavenged by CPTIO. NO was not detectable in control, isoprene-emitting (Fig. 3a) or isoprene-inhibited (Fig. 3b) leaves, in isoprene-emitting, ozonated leaves (Fig. 3c), or in isoprene-inhibited leaves on which NO was scavenged by CPTIO (Fig. 3d), but NO was found throughout the mesophyll of ozonated leaves whose emission of isoprene was prevented by fosmidomycin (Fig. 3e,f).

Figure 3.

Nitric oxide (NO) localization in mesophyll cells of Phragmites leaves. NO (bright yellow–green fluorescence; see arrows) was detected by epifluorescence light microscopy in leaf sections infused in the dark with 10 µm 4,5-diaminofluorescein diacetate for 3 h at 25°C. NO was not significantly revealed by the fluorescent probe in isoprene-emitting leaves (a), and in isoprene-inhibited leaves (b) that were not exposed to ozone. NO was also not detected in ozonated, isoprene-emitting leaves (c) and in ozonated, isoprene-inhibited leaves in which NO was scavenged by CPTIO (d). In the absence of NO-scavenging compounds, NO diffusely colonized the mesophyll cells of ozonated, isoprene-inhibited leaves, as shown in two different leaf sections (e, f). Chlorophyll autofluorescence is shown in red. Bars, 200 µm.


Isoprene emission is stimulated by oxidative stress and relieves ozone damage to the photosynthetic apparatus

The acute treatment with ozone stimulated the emission of isoprene in isoprene-emitting leaves. This is not the first time that a stimulation of isoprenoid emission is observed after a stress occurrence. Sharkey and Loreto (1993) observed that isoprene emission increased dramatically in plants recovering from water stress. It has also been shown that the emission of other isoprenoids increased strongly in ozonated leaves of Quercus ilex (Loreto et al., 2004). Our finding brings further support to the indication that the flux through the chloroplastic pathway of isoprenoid synthesis is stimulated by environmental stresses. However, isoprene emission might also increase because part of isoprene embedded into membranes to preserve their functions (Sharkey & Yeh, 2001) may also contribute to the emission once membranes become leaky under environmental stresses.

Isoprene emission is not regulated by stomatal opening (Monson & Fall, 1989) but is often associated with the amount of carbon fixed by photosynthesis (Sharkey & Yeh, 2001). In our ozonated leaves, both photosynthesis and stomatal conductance were reduced with respect to controls, indicating no relationship between increasing resistance to diffusion, or decreasing availability of photosynthetic intermediates, and isoprene emission. It remains to be ascertained whether the increasing emission reflects activation of different metabolic pathways of isoprene formation, or reflects higher activation of steps potentially regulating isoprene synthesis by the same chloroplastic pathway.

Isoprene-inhibited leaves suffered more ozone damage to photosynthesis than isoprene-emitting leaves, confirming our previous finding that isoprene may be involved in protection of the photosynthetic apparatus against oxidative stress (Loreto et al., 2001). It should be mentioned that isoprene protective action should not be exclusive of acute, perhaps unrealistic ozone concentration, because Loreto and Velikova (2001) reported this protection to occur also in leaves fumigated with 100 ng g−1 of ozone, a concentration normally reached in polluted urban and peri-urban areas. We brought evidence before that this protective action may be due to a direct reaction with ozone and ROS in the gas phase, or to an isoprene-mediated membrane stabilization (Loreto & Velikova, 2001). Our results showed that less H2O2 is formed, less activation of antioxidant enzymes and less membrane peroxidation occurs in ozonated, isoprene-emitting leaves than in leaves in which isoprene was inhibited by fosmidomycin. In addition to the two described mechanisms by which isoprene can protect against ozone, we now discuss the possibility that isoprene mediates NO and toxic ONOO accumulation and that this may also contribute to improve leaf resistance to ozone.

Isoprene quenches NO level

Our former observation that isoprene protects leaves from macroscopic damage in ozonated leaves (Loreto et al., 2001) was the reason we investigated whether isoprene could interact with NO. NO synthesis is enhanced in response to biotic and abiotic stress (Beligni & Lamattina, 2001; Gould et al., 2003). Many studies have shown that high concentrations of NO are phytotoxic, for instance through generation of toxic peroxynitrite, reduction of lipoxygenase activity (Beligni & Lamattina, 2001) or interaction with high levels of ROS, starting the cascade of signalling events leading to hypersensitive stress response and cellular death (Delledonne et al., 1998; Durner & Klessig, 1999). As in the experiment of Loreto et al. (2001), we have observed a necrotic damage significantly higher in isoprene-inhibited than in isoprene-emitting leaves 3 d after fumigation with ozone. We now show that, after the ozone treatment, NO is formed to a higher extent in leaves in which isoprene emission is inhibited by fosmidomycin than in isoprene-emitting leaves (Fig. 3). It is possible that the low formation of NO be an indirect consequence of the fact that isoprene reduces the oxidative stress at the primary metabolism level, as shown by our photosynthesis measurements. Photochemical energy is efficiently used to drive photosynthesis and less ROS-generating photoreduction of oxygen occurs in isoprene-emitting than in isoprene-inhibited leaves (Loreto & Velikova, 2001). When the oxidative burst is low, less NO is generated (Foissner et al., 2000). However, isoprene could also have effectively regulated NO levels in the leaf, probably reacting with NO radicals. Interestingly, isoprene emission seems to be inversely correlated with NO production. Both in controls and ozonated leaves, isoprene emission was highest when NO formation was inhibited by CPTIO feeding, and lowest when NO formation was enhanced by SNP feeding (Fig. 1). The difference was statistically significant only with P > 10%, but it also suggests a mutual interaction between isoprene and NO.

Irrespective of the mechanism by which isoprene is able to quench NO (direct scavenging of NO or ONOO, or a lower oxidative burst making less NO), we speculate that isoprene may effectively modulate the NO-triggered hypersensitive response at the basis of apoptosis in stressed leaves (Delledonne et al., 2001), and that this may be at the basis of the reduced damage observed in ozonated leaves by Loreto et al. (2001).

Isoprene may modulate hypersensitive responses to stress

We have demonstrated that only in isoprene-inhibited, ozonated leaves does the NO concentration become high enough to make its localization possible with the fluorescent probe. In the absence of isoprene, high amounts of NO (this work) and H2O2 (Loreto & Velikova, 2001) are simultaneously generated in ozonated leaves, thus activating the hypersensitive response leading to apoptosis, as suggested by Delledonne et al. (2001). By avoiding the build-up of NO and of its reactive products to toxic concentration, isoprene might modulate the signalling pathway activated by NO–ROS interactions and eventually leading to cellular death.

The discovery that another volatile molecule can modulate NO formation makes even more complex the understanding of volatile signalling and action. NO–isoprene interaction may have important consequences in the induction of systemic acquired resistance consequent to biotic infection, as many plants synthesize stress-induced isoprenoids (Paré & Tumlinson, 1997). It may also indirectly modulate other biosynthetic pathways (e.g. the cascade guanylate cyclase – phenylalanine ammonia-lyase/salicylic acid) that are triggered by NO-induction following stress episodes, and, though not directly related with oxidative cellular apoptosis, activate defence genes (Durner & Klessig, 1999). It may regulate long-distance signalling inducing systemic acquired resistance through the reaction of NO and glutathione. It may also regulate developmental signals leading to ethylene formation and accelerated senescence under stress conditions (Beligni & Lamattina, 2001).


This work was supported by the Italian Ministry for Environment programme OZONIT, and by the EC–Marie Curie Research and Training Network ‘Ecological and physiological functions of biogenic isoprenoids and their impact on the environment (ISONET)’. Violeta Velikova was supported by the NATO Collaborative Linkage Grant LST.CLG.978838 and by a CNR-NATO Outreach fellowship and Francesco Loreto by the Ministero dell’Istruzione, dell’Università e della Ricerca, Fondo per gli Investimenti della Ricerca di Base, programme RBAU018FWP.