American Journal of Respiratory and Critical Care Medicine

Direct measurements of endogenous nitric oxide (NO) release is of great interest but difficult to perform in vivo. We hypothesized that endogenous NO release from vasoactive substances would be detectable in exhaled air. Exhaled NO was measured after intravenous injections of various endothelium-dependent and endothelium-independent vasoactive drugs, in anesthetized pigs and humans. In pigs, a dose-dependent release of exhaled NO was observed for acetylcholine (ACh), bradykinin, substance P, endothelin (ET)-1, and nitroglycerine. Each compound had an individual and highly reproducible release pattern. Bradykinin-induced NO release was enhanced by angiotensin converting enzyme inhibition. ET receptor antagonism markedly reduced the response in exhaled NO to ET-1, whereas atropin abolished the NO response to ACh. NO synthase inhibition abolished basal levels of exhaled NO as well as the responses in exhaled NO to all compounds except nitroglycerine. In humans, ACh evoked a dose-dependent increase of NO levels in exhaled air. NO release by endogenous vasoactive agonists can be measured online in the exhaled air of pigs and humans. These novel findings may be useful when characterizing NO release from compounds that interfere with NO synthesis or drugs that act as donors of NO. Moreover, the possibility of using exhaled NO as an indicator of pulmonary endothelial dysfunction merits further studies.

There is a great interest in nitric oxide (NO) biology, but there are difficulties in performing direct in vivo measurements of this pluripotent gas. Most sensors are not sensitive enough and rapid to measure the very small quantities of NO that are transiently released. Moreover, the half-life of free NO in blood is extremely short because of its rapid reaction with proteins, e.g., hemoglobin (1). However, pulmonary NO production represents a unique exception. Because NO is a tiny lipid-soluble molecule that easily crosses biological membranes, this gas can escape into the airway lumen. NO in the gaseous phase is fairly stable, and it is measurable directly by chemiluminescence in exhaled air (2, 3). This allows for direct and noninvasive measurements of NO in different pulmonary disorders affecting the airways and/or the pulmonary vascular endothelium.

It has become evident that the vascular endothelium is not merely a passive barrier between the blood and surrounding tissues but in fact a highly active organ possessing many secretory, metabolic, and immunologic functions (4). During the last 2 decades there has been much focus on NO and endothelial function (for review see Reference 5). The capacity of the endothelium to produce NO in response to physiologic or chemical stimulants is crucial for the integrity of this major organ.

Many important vasoactive endogenous compounds like prostacyclin, NO, endothelin (ET), angiotensin II, endothelium-derived hyperpolarizing factor, free radicals, and bradykinin (BK) are formed in endothelial cells to control the functions of vascular smooth muscle cells and of circulating blood cells. Many vasodilators, e.g., acetylcholine (ACh) and BK, are known to stimulate endothelial NO release. The pulmonary endothelium represents a very large surface area, the major part of which lies in close contact with the alveoli.

Infusion of the NO donor glyceryl trinitrate (GTN) increases exhaled NO both in humans and animal models (68). In vitro experiments have shown the possibility of measuring ACh-induced NO release from the lung (9).

We hypothesized that endogenous endothelium–dependent vasoactive substances would give rise to a detectable release of NO in exhaled air. In a pig model, we characterized exhaled NO in relation to dosage, vascular responses, and effects of inhibitors. Moreover, using the same technique, we also studied the possibility of monitoring drug-induced endogenous release of NO in humans.

The Ethics Committees for animal and human experiments, Stockholm, Sweden, approved the experimental protocols for this study.

Anesthesia and Surgical Preparation

Domestic country breed pigs (n = 14) of both sexes, weighing 22 to 26 kg, were fasted overnight with free access to water. The pigs were premedicated with an intramuscular injection of ketamine (20 mg/kg), and anesthesia was induced by sodium pentobarbital (12–16 mg/kg) intravenously and maintained by a continuous intravenous infusion of midazolam (0.2 mg/kg/hour) and fentanyl (10 μg/kg/hour). The anesthetic level was evaluated by pain stimuli to the forehoof with a forceps. Additional doses of midazolam or fentanyl were given when needed. After tracheotomy, the animals were mechanically ventilated with a gas mixture of oxygen in air (volume-controlled V̇, FiO2, 0.30; Servo 900 ventilator; Siemens Elema, Umeå, Sweden). The respiratory frequency was set to 16 to 18 breaths/minute, and the V̇e was set to achieve similar Vts in each pig (300 ml). With this respiratory setting, maximal and mean expiratory flows were 392 ± 20 ml/second and 137 ± 19 ml/second, respectively. Body temperature was maintained at 38 to 39°C with a heated pad. A balloon-tipped pulmonary artery catheter was inserted under pressure guidance via a femoral vein to a position in the pulmonary artery for measurement of mean pulmonary arterial pressure (P̅p̅a̅). For measurement of mean arterial blood pressure (MAP), a catheter was introduced into the abdominal aorta via a femoral artery. A continuous infusion of isotonic saline with glucose (2.5 mg/ml) at a rate of 10 ml/kg/hour was maintained throughout the experiment via a catheter inserted in a femoral vein with its tip in the caval vein. A midline laparatomy was performed and a catheter was placed in the urinary bladder to collect urine. A left-sided thoracotomy was performed, and an ultrasonic flow probe (Transonic Systems, Ithaca, NY) was placed around the pulmonary artery for continuous registration of Q̇. MAP, Q̇, and Ppa were continuously recorded on a Grass polygraph (model 7B). After surgery, the animals were placed in the right lateral position and were allowed a 1-hour resting period before initiating the experimental protocol.

Measurements of Exhaled NO

Online measurements of NO were made with a chemiluminescence technique. To avoid contamination of NO in the inhaled air, the compressed wall air supplying the ventilator was passed through a charcoal filter. A linear pneumotachymeter (Hans Rudolph Inc., Kansas City, MO) was placed on the endotracheal tube for measurements of inhaled and exhaled flow rates. A portion of the inhaled and exhaled air (150 ml/min) was continuously sampled from a small catheter placed in the trachea into the NO-analyzing system (Exhaled Breath Analyzer; Aerocrine AB, Stockholm, Sweden). The signal output from the NO analyzer and the pneumotachymeter were connected to a computer-based system yielding instant onscreen display of flow, pressure, and NO concentration, and data were saved for postexperimental analysis. The detection limit for NO in this system is less than one part per billion (ppb). The NO analyzer was calibrated before each experiment with NO-free air and known concentrations of NO in nitrogen (AGA AB, Lidingö, Sweden).

Experimental Protocol

After the stabilization period, a series of vasoactive substances, terbutaline (4.5–15–45 nmol/kg), substance P (SP, 7.5–25–75 pmol/kg), prostacyclin (1–3–10 nmol/kg), papaverine (0.1–1–10 μmol/kg), GTN (4–12–40 nmol/kg), sodium nitroprusside (0.2–2–20 nmol/kg), ET-1 (1.5–15–150 pmol/kg) BK (0.3–1–3 nmol/kg), ACh (5–15–50 nmol/kg), and phenylephrine (15 nmol/kg), was given as intravenous bolus injections in the caval vein, in random order. In addition, repeated doses of ACh (15 nmol/kg) were given to each pig during the course of the experiments to assure that the responses were not subject to any spontaneous alterations concerning vascular reactivity and levels of exhaled NO. Single doses of selected substances were repeated after administration of atropin (0.5 mg/kg intravenous), the ACE-inhibitor captopril (1 mg/kg, intravenous), the ET receptor antagonist tezosentan (3 mg/kg, intravenous), the NOS inhibitor Nω-nitro-l-arginine methyl ester (L-NAME) (10 mg/kg intravenous) or its inactive enantiomer Nω-nitro-d-arginine methyl ester (10 mg/kg, intravenous). In a separate series of pigs (n = 4), ACh at doses between 0.15 and 50 nmol/kg were given before and after a higher dose of L-NAME (30 mg/kg). The doses of vasoactive substances were chosen on the basis of initial pilot experiments. Substances eliciting measurable levels of exhaled NO were given at doses just below and within a range where such release could be detected. Substances, on which no measurable exhaled NO levels could be revealed in pilot experiments, were given at doses exerting comparable vascular responses as those of ACh. Measurements of exhaled NO were timed to each injection of the different substances and were terminated when the response returned to baseline levels. A new bolus injection was made only when both the vascular and the exhaled NO response had returned to baseline.

Human Study

The subjects involved gave informed consent to the study. In six male patients (age 56–80 years, weight 68–99 kg) before coronary artery bypass graft operation, exhaled NO was measured on administering intravenous injections of ACh and GTN. Anesthesia was induced by midazolam (30–60 μg/kg, intravenous), fentanyl (6–15 μg/kg, intravenous), propofol (0.2–0.8 mg/kg, intravenous), and atracurium (0.5–0.7 mg/kg, intravenous) and maintained by a continuous infusion of propofol (4–8 mg/kg/hour, intravenous). The patients were mechanically ventilated (volume-controlled V̇, 20 breaths/minute, Vt 3.5–4 ml/kg, FiO2 0.5). For measurement of exhaled NO, a suction catheter (Mülly, Maersk, Denmark) was passed through the endotracheal tube. Gas from the distal orifice of the endotracheal tube was aspirated at a sampling flow of 300 ml/min. Exhaled NO, flow, and pressure were measured in the same way as in the pigs, and an NO scrubber (charcoal filter) was used to obtain zero NO in the inhaled gases. A balloon-tipped pulmonary artery catheter with thermistors for continuous Q̇ measurements was placed under pressure guidance in the pulmonary artery. A cannula was placed in the left radial artery for measurements of blood pressure, and continuous ECG tracing was used to monitor heart frequency.

After a stabilization period and measurements of baseline parameters, intravenous injections of ACh (17–210 nmol/kg) and GTN (6 nmol/kg) were given via the proximal port of the pulmonary artery catheter into the superior caval vein. Cardiovascular parameters and exhaled NO were measured at each dose of ACh and GTN, and values were allowed to return to baseline between each injection.


In humans, pulmonary vascular resistance index (PVRI) was calculated as ([P̅p̅a̅ − pulmonary capillary wedge pressure]/cardiac index) × 80. Cardiac index was calculated as Q̇/body surface area. In the pig, a simplified calculation of pulmonary vascular resistance (PVR) was made (Ppa/Q̇) because measurements of pulmonary capillary wedge pressure by itself led to increased levels of exhaled NO. Responses in exhaled NO are described as peak exhaled NO (ppb) and total release (area under curve). Total release was calculated as the total area under the curve for all exhalations during the response after the bolus injections and expressed as ppb × second. Comparison in relative potency between substances with regard to release of NO in exhaled air was calculated as the total release divided by the molar dose of each substance where the potency for ACh was set as 1. Cardiovascular responses are expressed as percent change from basal value. Data in the text are given as mean ± SEM, and statistical significance was calculated with the multiple analysis of variance followed by the post-test of Tukey or with the Student's t test (paired samples) where applicable. Statistical significance was considered when p values were less than 0.05.


Ketamine (Parke-Davis, Morris Plains, NJ), sodium pentobarbitone (Apoteksbolaget, Apoteket Labs., Umeå, Sweden), atropine and sodium heparin (KabiVitrum, Stockholm, Sweden), atracurium besylate (GlaxoSmithKline, Mölndal, Sweden), fentanyl (Pharmalink, Stockholm, Sweden), midazolam (Roche, Stockholm, Sweden), norepinephrine (Apoteket Labs., Umeå, Sweden), propofol (AstraZeneca, Södertälje, Sweden), GTN (TIKA, Lund, Sweden), sodium nitroprusside (Abbott Labs., Chicago, IL), terbutaline sulfate (Draco, Lund, Sweden), SP (Bachem, Bubendorf, Switzerland), ACh chloride (Clinalfa, Läufelfingen, Switzerland), atropine sulfate, BK acetate, captopril, ET-1, papaverine hydrochloride, phenylephrine hydrochloride, sodium prostacyclin, Nω-nitro-d-arginine methyl ester and L-NAME (Sigma, St. Louis, MO), tezosentan (Ro 61-0612; [5-isopropyl-pyridine-2-sulphonic acid 6-(2-hydroxy-ethoxy)-5-(2-methoxy-phenoxy)-2-(2–1H-tetrazol-5-yl-pyridin-4-yl)-pyrimidin-4-ylamide]) (Actelion, Allschwil, Switzerland).

Vascular Responses in the Pig

Basal MAP, Ppa, heart rate (HR), and Q̇ were 79 ± 7 mm Hg, 21 ± 1 mm Hg, 95 ± 8 beats/minute, and 3.3 ± 0.2 L/minute, respectively (n = 14). ACh, BK, SP, terbutaline, papaverine, prostacyclin, and GTN caused dose-dependent reduction of PVR and MAP and increase in Q̇. Sodium nitroprusside decreased PVR and MAP without significantly altering Q̇. ET-1 caused dose-dependent elevation of MAP with only marginal effects on Q̇ and PVR. Vascular responses observed at the highest dose of each substance are given in Table 1

TABLE 1. Vascular responses to various vasoactive substances in the anesthetized pig



Acetylcholine, 50 nmol/kg−55 ± 3+40 ± 7−21 ± 3
Bradykinin, 3 nmol/kg−69 ± 3+50 ± 10−36 ± 2
Substance P, 75 pmol/kg−50 ± 5+35 ± 5−29 ± 5
Endothelin-1, 150 pmol/kg+27 ± 2NSNS
Terbutaline, 45 nmol/kg−44 ± 3+52 ± 4−36 ± 2
Papaverine, 10 μmol/kg−54 ± 4+20 ± 3−13 ± 3
Prostacyclin, 10 nmol/kg−60 ± 13+34 ± 2−23 ± 13
Sodium nitroprusside, 20 nmol/kg−63 ± 3NS−18 ± 12
Glyceryltrinitrate, 40 nmol/kg−32 ± 2+15 ± 7−22 ± 7
Phenylephrine, 15 nmol/kg
+60 ± 9
−23 ± 5
+42 ± 12

Definition of abbreviations: MAP = mean arterial pressure; NS = not significant; PVR = pulmonary vascular resistance.

Values expressed as percent change from basal value.

Significance was set at p < 0.05, n = 4–6.

. Phenylephrine increased PVR and MAP, while lowering Q̇, at the single dose given (Table 1).

Captopril (1 mg/kg) lowered MAP (by 22 ± 7%) with only marginal effects on other parameters per se. Captopril markedly enhanced and prolonged the vascular response to BK (1 nmol/kg, MAP reduced by 60 ± 5% vs. control 47 ± 4%, the duration of this response was 430 ± 20 second vs. control 90 ± 10 second, both p < 0.05 vs. control, n = 5). Vascular responses to ACh (15 nmol/kg) and GTN (40 nmol/kg) were not affected by captopril (data not shown). Atropine (0.5 mg/kg) elevated HR and Q̇ (by 14 ± 5% and 12 ± 5%, respectively) without significant effects on other basal parameters per se. Atropine completely abolished the vascular response to ACh (50 nmol/kg, n = 4) but did not affect the response to BK (3 nmol/kg, e.g., MAP reduced by 62 ± 5% vs. control 67 ± 2%, not significant n = 4) and GTN (40 nmol/kg, data not shown). Tezosentan (3 mg/kg) lowered PVR and MAP (both by 16 ± 5%) and elevated HR and Q̇ (by 10 ± 3% and 9 ± 2%, respectively) per se. Tezosentan largely abolished the vascular response to ET-1 (150 pmol/kg, n = 4) but did not affect responses to ACh (15 nmol/kg) and BK (3 nmol/kg) (data not shown).

L-NAME (10 mg/kg) elevated PVR and MAP (by 75 ± 16% and 28 ± 7%, respectively) while lowering HR and Q̇ (by 8 ± 1% and 28 ± 7%, respectively) per se. L-NAME slightly inhibited the response to ACh (15 nmol/kg), whereby MAP was reduced by 29 ± 5% vs. control 35 ± 4% (p < 0.05, n = 5). The response to the higher dose of ACh (50 nmol/kg), however, was not significantly affected (data not shown). The response to BK (3 nmol/kg) was also only slightly affected by L-NAME (MAP reduced by 60 ± 4% vs. control 68 ± 4%, p < 0.05, n = 5), whereas the response to SP (75 pmol/kg) was not (data not shown). The effects of ET-1 (150 pmol/kg) were markedly altered after L-NAME (MAP increased by 45 ± 10% vs. control 27 ± 2%, Q̇ reduced by 45 ± 4% vs. control 4 ± 6%, PVR increased by 144 ± 13% vs. control 32 ± 28%, each p < 0.05 vs. control, n = 5). Furthermore, after L-NAME, the response to GTN (40 nmol/kg) was slightly augmented (MAP reduced by 37 ± 4% vs. control 32 ± 2%, Q̇ increased by 35 ± 8% vs. control 15 ± 9%, PVR reduced by 34 ± 5% vs. control 22 ± 7%, each p < 0.05 vs. control, n = 5). Nω-nitro-d-arginine methyl ester (10 mg/kg, n = 3) had no effects per se and did not affect the vascular response to ACh (15 nmol/kg) (data not shown). In a separate series of pigs, L-NAME (30 mg/kg) abolished vascular responses to ACh at doses between 0.15 to 0.5 nmol/kg and exerted significant inhibition of vascular responses to ACh at doses between 1.5 to 50 nmol/kg (data not shown).

Exhaled NO in the Pig

Inhaled NO was always below 1 ppb due to the charcoal filter used. Baseline exhaled NO was 2.0 ± 0.3 ppb and did not change during the course of the experiments.

There was a dose-dependent increase in exhaled NO by the endothelium-dependent vasodilators ACh (Figure 1)

, BK, and SP, as well as by ET-1 (Table 2)

TABLE 2. Peak levels of exhaled nitric oxide in response to increasing intravenous doses of different vasoactive compounds in the anesthetized pig



Glyceryl Trinitrate
NO (ppb)
NO (ppb)
NO (ppb)
52.9 ± 0.312.2 ± 0.144.2 ± 0.5
154.8 ± 0.432.6 ± 0.3129.5 ± 1
11.8 ± 0.4
2.6 ± 0.1
18.9 ± 1.7
Sodium Nitroprusside
NO (ppb)
NO (ppb)
NO (ppb)
0.34.1 ± 0.50.12 ± 0.4 0.23 ± 0.5
14.4 ± 0.512.6 ± 0.523.8 ± 0.7
5.6 ± 0.7
2.9 ± 0.3
13.7 ± 3.3
Substance P
NO (ppb)
NO (ppb)
NO (ppb)
7.52.5 ± 14.51.9 ± ± 0.5
252.8 ± 0.1152.4 ± 0.2156.6 ± 1.5
5.5 ± 0.7
2.3 ± 0.5
14.2 ± 1.3

Definition of abbreviation: NO = nitric oxide.

Mean basal NO levels were 2.0 ± 0.3 ppb.

. This was also seen with the endothelium-independent substances GTN and sodium nitroprusside (Table 2). In contrast, the vasodilators papaverine, terbutaline, and prostacyclin did not evoke any significant release of NO in exhaled air, as was the case also with the vasoconstrictor phenylephrine (Figure 2 , Table 2). The agents showed a varied profile in their release of NO in exhaled air. ET-1 gave by far the most prolonged release of NO in exhaled air, whereas the response to ACh was more rapid and of quite short duration (Figure 2). The peak exhaled NO levels for the substances are shown in Table 2. On a molar basis, the order of potency in release of NO in exhaled air of the administered endothelium-dependent compounds was ET-1 (3,914) more than SP (527) more than BK (10.4) more than ACh (1). In this context, the potency of GTN was 2.9 (Table 3)

TABLE 3. The ability of various compounds to generate nitric oxide in exhaled air in the anesthetized pig


Relative NO Release/mol
Glyceryl trinitrate2.9 
Substance P527 

For definition of abbreviation see Table 2.

The relative potency between substances was calculated as the total release divided by the molar dose of each substance where the potency for acetylcholine was set as 1.

. The NO release caused by ACh (15 nmol/kg) in the pig was highly reproducible and did not change during the course of the experiments.

Atropine (0.5 mg/kg, n = 4) abolished the NO response to ACh (50 nmol/kg) (Figure 3)

, whereas the NO release on administering BK (3 nmol/kg) and GTN (40 nmol/kg) was unaffected (data not shown). Tezosentan (3 mg/kg) inhibited (by 80 ± 17%, p < 0.05, n = 4) exhaled NO induced by ET-1 (150 pmol/kg) (Figure 3) but did not affect responses to ACh (15 nmol/kg) and BK (3 nmol/kg) (data not shown). Captopril (1 mg/kg) increased (by 75 ± 25%, p < 0.05, n = 5) the NO release detected on injecting BK (1 nmol/kg) (Figure 3) without affecting the NO release on injecting ACh (15 nmol/kg) and GTN (40 nmol/kg) (data not shown). Neither atropine, tezosentan, or captopril had any effect on basal levels of exhaled NO. In contrast, L-NAME (10 mg/kg, n = 5) led to a total loss of detectable NO in exhaled air within 1 minute, and NO release was abolished for all endothelium-dependent vasodilators tested. However, the response to GTN was unchanged after L-NAME (data not shown). Nω-nitro-d-arginine methyl ester (10 mg/kg, n = 3) did not affect either basal or ACh-evoked NO release (data not shown).

Human Study

Inhaled NO was below 1 ppb, and basal exhaled NO levels were 2.9 ± 0.6 ppb. Baseline exhaled NO was unchanged during the course of the experiments. Basal MAP, HR, Q̇, and PVRI were 90 ± 6 mm Hg, 42 ± 2 beats/minute, 3.8 ± 0.4 L/minute, and 350 ± 50 dyne second cm−5 m2. ACh (17–210 nmol/kg, intravenous) injections dose-dependently increased NO in exhaled air (Figures 4 and 5)

. ACh (140 nmol/kg) decreased PVRI by a moderate 28 ± 13% due to small changes in Ppa and pulmonary capillary wedge pressure. No clear dose–response relationship could be observed for ACh on PVRI. MAP, HR, and Q̇ were not, or only marginally, affected by ACh at any dose given. In one patient, GTN (6 nmol/kg, intravenous) decreased MAP and PVRI by 20 and 13%, respectively, whereas HR was elevated by 19% and Q̇ was unaffected. In the following patients, because of these negative circulatory effects, an infusion of norepinephrine (50–200 ng/kg/minute, intravenous) was administered simultaneously as GTN was given, thereby precluding relevant measurements of the cardiovascular effects of GTN per se. GTN (6 nmol/kg) elevated levels of exhaled NO by 2.3 ± 0.2 ppb (n = 4).

We show here that intravenous administration of several endothelium-dependent vasoactive substances cause dose-dependent release of exhaled NO in a highly reproducible manner. There was a marked difference between the compounds with regard to the profile and total release of NO. These findings demonstrate the possibility of using exhaled measurements to detect endogenous NO release induced by vasoactive substances. The inhibitory effects of atropine and tezosentan clearly demonstrated receptor-mediated stimulation of NO release for ACh and ET, respectively. Furthermore, the inhibition of L-NAME showed the enzymatic origin of the NO release caused by endothelium-dependent vasodilators ACh, BK, and SP, as well as ET, Moreover, by interfering with the degradation of BK (with captopril), both exhaled NO and vascular responses were enhanced in parallel.

It has been very difficult to directly measure the minute amounts of NO transiently released by the endothelium on stimulation with, e.g., ACh. Therefore, one is most often confined to indirect measurements of NO reaction products such as nitrite, nitrate, and S-nitrosothiols. However, such measurements are less sensitive and unable to display the rapid dynamics in NO release. In addition, analysis has to be done ex vivo and is time consuming. Direct measurement of NO in the circulation has been described by Malinski and coworkers using intravascular chemical sensors (10). Although somewhat invasive, this method may be useful in certain scenarios but is to date not applicable in the pulmonary vasculature.

The anatomic origin of exhaled NO in this study was not investigated. Several methods based on mathematic two-compartment models of the airways and alveoli have been used in awake humans to try to locate the origin of exhaled NO. They are based on controlled exhalations with plateau NO levels at different flow rates. In humans, NO released from the airway wall is highly flow-dependent (11, 12), whereas NO from the alveolar region is not (13). These techniques are difficult to perform in anesthetized, nonparticipating pigs and humans. Nevertheless, there are some other indications that a major part of the NO measured here is in fact derived from a distal source of the airways (alveolar epithelium) or the pulmonary vascular endothelium. The animals were tracheotomized, which excludes any contribution from the upper airways (14). Also, as mentioned in methods, inflation of the pulmonary artery catheter balloon led to an instant increase in exhaled NO. This could have been the effect of a stopped blood flow in a part of the lung leading to diminished scavenging of endothelial or alveolar NO in that pulmonary region. Finally, if the NO release on administering ACh was derived from the airway wall, one would expect that other muscarinic receptor agonists, e.g., metacholine, would also result in increased NO when administered locally in the airways. However, this is not the case as has been shown in bronchial provocation studies in humans (15). Further studies, e.g., using multiple flow techniques, could probably help to better locate the origin of exhaled NO after agonist stimulation.

Interestingly, ET-1 was by far the most potent mediator in increasing exhaled NO compared with both endothelium-dependent endogenous vasodilators as well as exogenous NO donors. ET-1 is considered to be one of the most potent endogenous vasoconstrictors in humans acting on at least two different receptor subtypes. Activation of the ETA receptor subtype located on the vascular smooth muscle leads to prolonged vasoconstriction, whereas ETB receptor subtype activation on the vascular endothelium leads to vasodilation through release of NO and/or prostacyclin (16, 17). There are also ETB receptors on the vascular smooth muscle that induce vasoconstriction. Endogenous ET is considered to act in a paracrine fashion on the underlying vascular smooth muscle. However, after bolus injections of exogenous ET, the response is often a short initial vasodilation followed by a prolonged vasoconstriction. Clearance of circulating ET is mediated by the ETB receptors in the lung (16). In the pulmonary vasculature, the overall net effects of ET-1 depend on dose, pre-existing vascular tone, and the distribution of receptor subtypes. In the pigs, ET-1 did not evoke any major pulmonary vascular changes independent of dose because the vasoconstrictory properties of this peptide were effectively counteracted by the concomitant release of NO. However, L-NAME unmasked this ETB receptor–mediated counteraction, and a marked pulmonary vasoconstriction was evident. It is fascinating to note that behind the vascular response to this extremely potent vasoconstrictor, NO is released in amounts greatly outnumbering that of all the pure vasodilators tested here (including GTN).

L-NAME abolished the increases in exhaled NO observed at endothelium-dependent vasoactive substances. In contrast, it became evident that L-NAME, with the exception of the ET response, exerted, at most, very marginal inhibition of the vascular responses to the endothelium-dependent vasodilators ACh, BK, and SP. These latter substances, however, were given at rather high doses (high enough to yield measurable levels of NO in exhaled air). The moderate inhibitory effect exerted by L-NAME on the vascular responses to these doses was likely a dose-related phenomenon. Thus, a higher dose of L-NAME significantly inhibited vascular responses to these high doses of ACh (that yielded measurable increases in exhaled NO before L-NAME treatment) as well as abolished vascular responses to lower doses of ACh. Thus, it seems that inhibition of NO-synthase in vivo with L-NAME is not effective enough to completely antagonize stimulated NO production on administering high doses of endothelium-dependent vasodilators. Because L-NAME abolished the exhaled levels of NO on administering these high doses of endothelium-dependent vasodilators, this detectable NO release may represent a small spillover from a much larger release within the tissue, the rest of which is possibly buffered by hemoglobin or inactivated in other ways. Interestingly, L-NAME markedly altered the vascular response to ET. Thus, when NO counteracts vasoconstriction, as for ET, the involvement of NO seems crucial for the outcome of the vascular response. That L-NAME in comparison affected the ET response more than the endothelium-dependent vasodilators may also be related to the doses given. Evidently, ET was a much more potent stimulator of NO release than the other substance studied, and it also exerted quite a moderate vascular response at the dose given, thus becoming more susceptible to NOS inhibition. In agreement with previous in vivo studies, the effects of GTN were slightly augmented after treatment with L-NAME. In a study by Moncada and coworkers, this was explained by a removal of the basal NO-mediated vasodilator tone leading to hypersensitivity to nitrovasodilators at the level of the soluble gunalylyl cyclase (18). Accordingly, the actual release of NO, as measured in exhaled air, to GTN was not affected after L-NAME in the present study.

There was an obvious difference between pigs and humans in the response to ACh in this study. First, approximately 10-fold higher doses were needed in humans to give a similar release in exhaled NO and reduction in PVRI. Second, no systemic response (decrease in MAP) was seen in humans, whereas a marked fall in MAP was evident at all the three doses in the pig. This is most likely due to differences in efficacy of the circulating ACh esterases between the two species. Indeed, the plasma esterase activity in the pig seems lower than in many other species (19). Furthermore, plasma esterase activity in pigs increases with age (20), and it should be noted that in the present study pigs at ages between 2 and 3 months were studied. In humans, breakdown of circulating ACh is highly efficient, and the site of injection strongly determinates the effect (21). Moreover, there was no effect on heart rate in the patients, which speaks against any remaining ACh entering the left side of the heart. This is satisfying because ACh reaching the coronary circulation might cause vasoconstriction in patients with coronary artherosclerosis and damaged endothelium (22). Therefore, ACh seems to be a suitable agonist to induce endothelium-dependent, NO-releasing vasodilation in the pulmonary circulation without causing negative effects on cardiac performance and the systemic circulation.

Much effort has been put to try to evaluate the function of the vascular endothelium to monitor treatment or predict future cardiovascular disease. Several tests have been proposed such as forearm blood flow response and coronary artery diameter response to ACh and brachial artery flow–mediated dilatory response to postischemic hyperemia (23). Also, the function of the pulmonary vascular endothelium has been studied but with more invasive techniques that involve measurements of pulmonary vasoreactivity to various agonists (24). Clearly NO is important in endothelial dysfunction, but the exact mechanism is yet to be elucidated. A decreased synthesis or an increased breakdown of NO has been suggested. In addition, impaired cellular responses to NO have also been discussed (25).

If future studies demonstrate that, e.g., ACh-evoked exhaled NO release truly is derived from the pulmonary vascular endothelium, an exhaled NO test could possibly be used to reveal endothelial dysfunction. Naturally, the nature of the underlying defect in the NO pathway in endothelial dysfunction will determine this. Endothelial dysfunction is present at an early stage in many disorders of the cardiovascular system, including hypertension, cardiac failure, hypercholesterolemia, sepsis, and diabetes (25). However, whether the pulmonary endothelium is also affected in any of these conditions is less certain but of interest because the present methodology is evidently limited to the pulmonary circulation. If ACh-evoked release of exhaled NO reflects endothelial function, it may be of use in disorders affecting the pulmonary circulation. The pulmonary endothelium is clearly affected in acute lung injury, chronic obstructive pulmonary disease, acute respiratory distress syndrome, and sepsis. It will be interesting to study if the increase in exhaled NO evoked by, e.g., ACh is attenuated in such patients with pulmonary endothelial dysfunction. Nitroglycerine and nitroprusside were used in the present study as endothelium-independent NO releasing agents. They are obviously also needed as “positive controls” in situations where altered NO release on administering endothelium-dependent agonists may occur.

We conclude that endogenous NO release induced by vasoactive substances can be monitored online in exhaled air of pigs and humans. Furthermore, exhaled NO may be useful to characterize NO release from compounds that interfere with NO synthesis or drugs that act as donors of NO. Once the exact source of agonist-evoked release of NO in exhaled air is determined, the usefulness of this procedure to detect pulmonary endothelial dysfunction can be evaluated.

The authors thank Ms. M. Stensdotter for expert technical assistance.

1. Moncada S, Palmer RM, Higgs EA. Nitric oxide: physiology, pathophysiology, and pharmacology. Pharmacol Rev 1991;43:109–142.
2. Lundberg JO, Lundberg JM, Alving K, Weitzberg E. Nitric oxide and inflammation: the answer is blowing in the wind. Nat Med 1997;3:30–31.
3. Gustafsson LE, Leone AM, Persson MG, Wiklund NP, Moncada S. Endogenous nitric oxide is present in the exhaled air of rabbits, guinea pigs and humans. Biochem Biophys Res Commun 1991;181:852–857.
4. Cines DB, Pollak ES, Buck CA, Loscalzo J, Zimmerman GA, McEver RP, Pober JS, Wick TM, Konkle BA, Schwartz BS, et al. Endothelial cells in physiology and in the pathophysiology of vascular disorders. Blood 1998;91:3527–3561.
5. Vallance P, Chan N. Endothelial function and nitric oxide: clinical relevance. Heart 2001;85:342–350.
6. Agvald P, Adding LC, Gustafson LE, Persson MG. Nitric oxide generation, tachyphylaxis and cross-tachyphylaxis from nitrovasodilators in vivo. Eur J Pharmacol 1999;385:137–145.
7. Marczin N, Riedel B, Royston D, Yacoub M. Intravenous nitrate vasodilators and exhaled nitric oxide. Lancet 1997;349:1742.
8. Persson MG, Agvald P, Gustafsson LE. Detection of nitric oxide in exhaled air during administration of nitroglycerin in vivo. Br J Pharmacol 1994;111:825–828.
9. Cremona G, Higenbottam T, Takao M, Hall L, Bower EA. Exhaled nitric oxide in isolated pig lungs. J Appl Physiol 1995;78:59–63.
10. Malinski T, Mesaros S, Patton SR, Mesarosova A. Direct measurement of nitric oxide in the cardiovascular system. Physiol Res 1996;45:279–284.
11. Lundberg JO, Weitzberg E, Lundberg JM, Alving K. Nitric oxide in exhaled air. Eur Respir J 1996;9:2671–2680.
12. Silkoff PE, McClean PA, Slutsky AS, Furlott HG, Hoffstein E, Wakita S, Chapman KR, Szalai JP, Zamel N. Marked flow-dependence of exhaled nitric oxide using a new technique to exclude nasal nitric oxide. Am J Respir Crit Care Med 1997;155:260–267.
13. Jorres RA. Modelling the production of nitric oxide within the human airways. Eur Respir J 2000;16:555–560.
14. Tornberg DC, Marteus H, Schedin U, Alving K, Lundberg JO, Weitzberg E. Nasal and oral contribution to inhaled and exhaled nitric oxide: a study in tracheotomized patients. Eur Respir J 2002;19:859–864.
15. Garnier P, Fajac I, Dessanges JF, Dall'Ava-Santucci J, Lockhart A, Dinh-Xuan AT. Exhaled nitric oxide during acute changes of airways calibre in asthma. Eur Respir J 1996;9:1134–1138.
16. de Nucci G, Thomas R, D'Orleans-Juste P, Antunes E, Walder C, Warner TD, Vane JR. Pressor effects of circulating endothelin are limited by its removal in the pulmonary circulation and by the release of prostacyclin and endothelium-derived relaxing factor. Proc Natl Acad Sci USA 1988;85:9797–9800.
17. Yanagisawa M, Kurihara H, Kimura S, Tomobe Y, Kobayashi M, Mitsui Y, Yazaki Y, Goto K, Masaki T. A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature 1988;332:411–415.
18. Moncada S, Rees DD, Schulz R, Palmer RM. Development and mechanism of a specific supersensitivity to nitrovasodilators after inhibition of vascular nitric oxide synthesis in vivo. Proc Natl Acad Sci USA 1991;88:2166–2170.
19. Silvestri R. New techniques to measure blood cholinesterase activity in domesticated animals. Am J Vet Res 1977;38:659–662.
20. Murphy TM, Mitchell RW, Blake JS, Mack MM, Kelly EA, Munoz NM, Leff AR. Expression of airway contractile properties and acetylcholinesterase activity in swine. J Appl Physiol 1989;67:174–180.
21. Duff F, Greenfield ADM, Shepherd JT, Thomson ID. A quantative study of the response to acetylcholine and histamine of the blood vessels of the human hand and forearm. J Physiol 1953;120:160–170.
22. Ludmer PL, Selwyn AP, Shook TL, Wayne RR, Mudge GH, Alexander RW, Ganz P. Paradoxical vasoconstriction induced by acetylcholine in atherosclerotic coronary arteries. N Engl J Med 1986;315:1046–1051.
23. Monnink SH, van Haelst PL, van Boven AJ, Stroes ES, Tio RA, Plokker TW, Smit AJ, Veeger NJ, Crijns HJ, van Gilst WH. Endothelial dysfunction in patients with coronary artery disease: a comparison of three frequently reported tests. J Investig Med 2002;50:19–24.
24. Celermajer DS, Cullen S, Deanfield JE. Impairment of endothelium-dependent pulmonary artery relaxation in children with congenital heart disease and abnormal pulmonary hemodynamics. Circulation 1993;87:440–446.
25. Mombouli J-V, Vanhoutte PM. Endothelial dysfunction: from physiology to therapy. J Mol Cell Cardiol 1999;31:61–74.
Correspondence and requests for reprints should be addressed to Eddie Weitzberg, M.D., Ph.D., Department of Surgical Sciences, Karolinska Institute, S-17176 Stockholm, Sweden. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record