Ultracytochemical Localization of the NADPH-d Activity in the Human Nasal Respiratory Mucosa in Vasomotor Rhinitis



Objectives Description of the ultrastructural localization of nitric oxide synthase in the blood vessels of the nasal respiratory mucosa in patients with vasomotor rhinitis.

Study Design This research was conducted on seven patients—men and women, ages 20 to 45 years—suffering from vasomotor rhinitis and who had undergone surgical therapy for reduction of the inferior turbinates.

Methods To study the ultrastructural localization of nitric oxide synthase, NADPH-diaphorase cytochemistry was employed. Samples of the nasal mucosa were obtained from inferior turbinates.

Results The endothelial cells of the arterioles, capillaries , venules and cavernous sinuses revealed a distribution of the enzymatic activity similar to that found in unaffected subjects. A strong enzymatic activity was recognized in the smooth muscle cells of the cavernous sinuses. The smooth muscle cells of arterioles and venules were generally found to be negative to enzymatic reaction.

Conclusions This study suggests that the vascular disorders of the vasomotor rhinitis depend, at least in part, from nitric oxide synthase induction in the smooth muscle cells of the cavernous sinuses.


Vasomotor rhinitis is a dysfunction of undecided etiology usually defined as a nonimmunological, noninfectious, chronic type of rhinitis, without nasal eosinophilia. 1–3 Vasomotor rhinitis is characterized by a profuse watery rhinorrhea and mucosal congestion due to vascular dilatation. Pharmacological treatment of vasomotor rhinitis is often frustrating. Poor response to vasoconstrictive sprays, decongestants, antihistamines and corticosteroids is generally found. Anticholinergic agents may be effective in reducing the severity and the duration of the rhinorrhea, but their use does not result in an improvement of the nasal congestion. 3 Therefore, dilatation of the vessels appears as the more resistant symptom of vasomotor rhinitis. Blood vessels of the nasal respiratory mucosa (NRM) consist of resistance vessels (arteries, arterioles), capillaries, and capacitance vessels (cavernous sinuses, veins). 4 The cavernous sinuses form a venous erectile tissue and are the most distinctive vascular structure of the NRM. These vessels may be filled with blood in response to a variety of physical and chemical stimuli, thus causing congestion of the NRM. 5–10 The persistent mucosal swelling observed in vasomotor rhinitis is the consequence of the cavernous sinus dilatation. 1–3 Recently nitric oxide (NO) has been proposed as a mediator of important functions in the nose. NO may cause relaxation of vascular smooth muscle cells (VSMC) and probably modulates glandular secretion and ciliary beat frequency in the respiratory epithelium. 10–17 In addition, evidence indicates that NO is the mediator of the effector arm of reflexes increasing vascular permeability in the nose. 18,19 NO is synthesized from the amino acid L-arginine by a group of isoenzymes collectively named nitric oxide synthase (NOS). 20–22 At present three isoforms of NOS have been identified: two of these, neuronal isoform (n-NOS) and endothelial isoform (e-NOS), are Ca2+-dependent and constitutively expressed. The third species of the enzyme (i-NOS) is a Ca2+-independent inducible isoform, virtually expressed by any cell type when adequately stimulated with cytokines, endotoxins, and other agents. 22 Unlike constitutive isoforms, i-NOS has a continuous full catalytic activity and may produce large quantities of NO until substrates become limiting. 21,22 The enzyme activity of NOS is dependent on the oxidation of its co-factor, reduced NADPH. Thus, NADPH-diaphorase (NADPH-d) histochemistry has been used as marker for NOS in aldehyde-fixed tissues. 23–27 In the NRM the NADPH-d activity was observed to co-localize with NOS. 11–13 NOS-containing nerve fibers were found in the NRM around blood vessels and submucosal glands. 11–14 Intraepithelial arborizations of NOS-immunoreactive fibers were also described. 12 Recently, the ultracytochemical distribution of NADPH-d enzymatic activity was described in the blood vessels of the normal human NRM. 17 Strong enzymatic activity was observed in the endothelium of the cavernous sinuses. The endothelial cells of the capillaries and venules were found to lack NADPH-d and the endothelium of the arterioles occasionally appeared labeled. These observations are indicative of a major role of the endothelial NO in the capacitance vessels compared with the resistance vessels. In patients with chronic rhinitis an increase in NOS immunoreactivity and in NADPH-d histochemical staining has been shown in the surface epithelium, glands, and blood vessels. 13 This finding has suggested NO to be involved in the mucosal edema and increased secretion associated with nasal diseases. However, the role of NO in vasomotor rhinitis has not been specifically studied or elucidated. The aim of this work was, therefore, to investigate, by means of NADPH-d ultracytochemistry, the NOS distribution in the endothelial cells, and VSMC of the NRM in patients affected by vasomotor rhinitis.


The study was carried out on seven patients (four men and three women; age range, 20–45 y; mean age, 33.2 ± 1.8 y) with clinical diagnosis of vasomotor rhinitis, who had undergone surgical therapy for reduction of the inferior turbinates, in absence of significant septal deviation. The patients were affected by the typical symptoms of vasomotor rhinitis and showed hypertrophy of the inferior turbinate significantly obstructing the nasal cavities. The evaluation of nasal patency was carried out by active anterior rhinomanometry. The nasal resistance at a pressure of 150 Pa was in all cases higher than 0.90 Pa/s per milliliter. All the patients showed negative nasal smear test for eosinophils, negative skin prick test, and negative radioallergosorbent test. None smoked, none had previous or on-going respiratory disease, and none had received therapy within 1 month before undergoing surgery. Samples of nasal mucosa were obtained from inferior turbinates and were immersed in a solution of 4% paraformaldehyde and 1% glutaraldehyde in cold 0.1 mol/L phosphate buffer, pH 7.2 for 2 hours. After a brief washing in buffer, the specimens were processed for NADPH-d histochemistry by the method of Scherer-Singler et al., 28 modified for electron microscopy by Wolf et al. 24 In brief, the specimens were immersed for 90 minutes at 37°C in a solution of 0.1 mol/L phosphate buffer, pH 8.0, containing 1.2 mmol/L β-NADPH and 1.2 mmol/L 2-(2′-benzothiazolyl)-5-styryl-3-(4′-phthalhydrazidyl) tetrazolium chloride. The reaction was stopped by rinsing the specimens in cold 0.1 mol/L phosphate buffer, pH 7.2. Specimens incubated with NADPH-free medium were utilized as controls. The specimens were postfixed in osmium tetroxide 1.3% in the same buffer for 1 hour, dehydrated in alcohol, and embedded in Epon 812, following the usual procedures. To establish the percentage of the endothelial cells and of the VSMC positive for the NADPH-d in the various blood vessels, serial ultrathin sections 100-nm thick were obtained and every third section was mounted on a single-hole Formvar-coated grid. In addition, the endothelial cells and the VSMC positive for the enzymatic reaction were counted in ultrathin sections taken from different levels of the specimens. The sections were double stained with saturated uranyl acetate in 50% alcohol and with lead citrate, then observed under a JEOL 100-SX electron microscope (JEOL USA, Inc., Peabody, MA).


Examination of NADPH-d activity in the endothelial cells of the NRM in patients affected by vasomotor rhinitis did not reveal significant differences compared with the endothelial cells of the normal NRM. 17 The enzymatic reaction was absent in the endothelium of the venules and capillaries. A small number of NADPH-d–positive endothelial cells (approximately 5%) were recognized in the arteriolar wall. As in normal NRM (Figs. 1 and 2), all the endothelial cells of the cavernous sinuses were intensely stained (Figs. 3 and 4). Labeling was deposited on the smooth endoplasmic reticulum and on the nuclear cistern (Fig. 4). Approximately 85% of the VSMC of the cavernous sinuses were found to be positive for the enzymatic reaction (Fig. 3). The reaction product was observed in the mitochondria (Fig. 5), in the nuclear cistern (Fig. 5), and in the cisternae of the smooth endoplasmic reticulum (Fig. 6). VSMC of venules and arterioles were generally NADPH-d–negative. With omission of NADPH from the incubation medium, all staining was abolished.

Figure Fig. 1..

Normal nasal respiratory mucosa. The enzymatic reaction is observed in the endothelium (arrows). The vascular smooth muscle cells (VSMC) are not labeled (arrowheads). Scale bar: 0.80 μm.

Figure Fig. 2..

Normal nasal respiratory mucosa. NADPH-d staining is shown in the endothelium. The enzymatic reaction is found in the cytoplasm (arrows) and in the nuclear cistern (arrowheads). Scale bar: 0.90 μm.

Figure Fig. 3..

Nasal respiratory mucosa in vasomotor rhinitis. Labeling is found in the endothelium (arrowheads) and in several VSMC (arrows). Scale bar: 0.82 μm.

Figure Fig. 4..

Nasal respiratory mucosa in vasomotor rhinitis. Electron micrograph showing an endothelial cell. The enzymatic reaction is found in the nuclear cistern (arrowheads) and in the smooth endoplasmic reticulum (arrows). Scale bar: 1.6 μm.

Figure Fig. 5..

Nasal respiratory mucosa in vasomotor rhinitis. Some positively stained mitochondria are shown (arrows). The nuclear cistern is also labeled (arrowhead). The small arrows point to unlabeled mitochondria. Scale bar: 2.2 μm.

Figure Fig. 6..

Nasal respiratory mucosa in vasomotor rhinitis. Section of a VSMC; the smooth endoplasmic reticulum is intensely labeled (arrows). N = nucleus. Scale bar: 2 μm.


The mechanisms by which the nasal blood vessels are controlled are still not completely known. The nasal blood vessels receive fibers from the sympathetic and the parasympathetic nervous systems. In normal subjects sympathetic stimulation results in a constriction of both resistance and capacitance vessels, with an increase in nasal patency. 9,10,29 Activation of the parasympathetic fibers decreased vascular resistance and increased vascular capacitance, causing a reduction of nasal patency. 9,10,29,30 These effects were atropine-resistant 29 and were reduced by further 80% by contemporarily administrating N-ωnitro-L-arginine (L-NNA), a NOS inhibitor. 30 This suggests that NO has a crucial role in the vascular control of normal NRM as noncholinergic parasympathetic neurotransmitter. When both sympathetic and parasympathetic nerves are stimulated, vasoconstriction occurs. 9,29 In addition, parasympathetic denervation does not significantly alter vascular resistance or vascular capacitance. 9,29 Therefore, it is believed that a predominant sympathetic control and a negligible parasympathetic tone probably exist in nasal blood vessels. 9,29 Vidian neurectomy has been sometimes performed in patients affected by vasomotor rhinitis. 3,31 This surgical procedure, which interrupts both sympathetic and parasympathetic fibers bound to NRM, was effective in dealing with the rhinorrhea associated with vasomotor rhinitis but resulted in little improvement of the nasal congestion. This suggests that NO produced by parasympathetic nerve terminals does not cause the cavernous sinus dilatation observed in vasomotor rhinitis, and neither does it support the hypothesis of increased parasympathetic reflexes in the genesis of the vascular disorders in vasomotor rhinitis. It has been speculated that the nasal congestion mechanism in vasomotor rhinitis could consist in a decrease of sympathetic influences rather than in the overactivity of parasympathetic fibers, 9 but precise evidence of this hypothesis is lacking. In NRM, trigeminal sensory neurons contain calcitonin gene–related peptide and other neuropeptides such as the tachykinins substance P and neurokinin A. 7,10,32–34 The activation of sensory neurons leads to the recruitment of vascular and secretory reflexes by releasing neuropeptides in the central nervous system. In addition, local reactions such as vasodilatation and increased secretory activity may result from nerve axon responses. 7,10,33,34 The importance of this functional mechanism in the NRM is a controversial topic. After administration of atropine plus a ganglion blocker, stimulation of sensory terminals with capsaicin caused only a slight increase in the superficial blood flow, 7,34 indicating that nerve axon responses probably account little for the vascular control of the NRM in normal conditions. At present it has not been clearly demonstrated if increased sensitivity of nociceptive neurons significantly contributes to vascular symptoms of vasomotor rhinitis via nerve axon responses.

In conclusion, it can be asserted that in vasomotor rhinitis increased glandular secretion is probably caused by hyperactive parasympathetic reflexes, whereas dilatation of the cavernous sinuses is of uncertain etiology. Our observations suggest one possible pathophysiological mechanism of this symptom. NADPH-d activity was generally absent in the VSMC of arterioles, venules, and cavernous sinuses of normal subjects. On the contrary, in patients suffering from vasomotor rhinitis an intense enzymatic reaction was present in the VSMC of the cavernous sinuses (Figs. 3, 5, and 6). As far as the endothelial cells are concerned, all examined vessels showed a NADPH-d activity similar to the ones observed in normal subjects. 17 This suggests that NO produced by endothelial cells does not have a direct role in vasomotor rhinitis. Presence of NOS activity in VSMC reflects the expression of i-NOS, the inducible form of NOS. Indeed, it has been observed that VSMC not only respond to endothelium-derived NO but may also synthesize their own NO by induction of i-NOS. 22,35,36 Thus when i-NOS induction in VSMC becomes excessive, the release of large amounts of NO may cause profound vasodilatation and hypotension, as occurs in sepsis-induced and cytokine-induced circulatory shock. 22,35,36 The data here reported suggest that the persistent relaxation of the cavernous sinuses found in vasomotor rhinitis is probably caused by a strong induction of i-NOS in their VSMC. The course of events in the induction of i-NOS is very complex and not entirely known. 22,35 At present it is not possible to clearly identify substances acting on VSMC of the cavernous sinuses in the course of vasomotor rhinitis. However, considering that vasomotor rhinitis consists in a nonimmunological, noninfectious type of rhinitis, it can be hypothesized that physiopathological conditions leading to NO overproduction in VSMC of the cavernous sinuses might be different from those occurring in VSMC in immune activation conditions, such as septic shock, where bacterial products and cytokines play the main role. However, the little ability of the exogenous vasoconstrictors to reduce the mucosal congestion in vasomotor rhinitis 3 is similar to the diffused hyporesponsivity to the same substances shown by vessels in septic shock. 36 Stimuli other than cytokines and bacterial products affecting i-NOS expression in VSMC might be represented by cAMP-elevating agents. 37–39 In fact, it has been observed that intracellular cAMP may regulate i-NOS gene expression in VSMC by stimulating i-NOS gene transcription and by decreasing the degradation rate of i-NOS mRNA. 37–39 VSMC of the cavernous sinuses have receptors for sensory neuropeptides acting as vasodilators by means of a cAMP-dependent mechanism, as receptors for the sensory neuropeptides calcitonin gene–related peptide and substance P. 10,34 Thus hyperactivity of sensory fibers in vasomotor rhinitis could cause increased secretory activity by means of activation of parasympathetic reflexes and, at the same time, by means of nerve axon responses, i-NOS induction in VSMC through a cAMP-dependent mechanism. An implication of this hypothesis is that neuropeptides do not induce NOS in normal sensory fiber activity. Actually, it has been observed that calcitonin gene–related peptide may induce i-NOS in the VSMC with an intense but not with a moderate activation of the cAMP effector pathway. 39 Unlike the VSMC of the cavernous sinuses, VSMC of the arterioles and venules, despite the presence of many receptors for sensory neuropeptides, 10,34 did not express NOS activity in vasomotor rhinitis. Functional relationships between endothelial cells and VSMC could explain this puzzling observation. Besides NO, vascular endothelial cells have been shown to produce a substance called endothelin-1, the most potent vasoconstrictor identified to date. 40 Evidences suggest that endothelin-1 and NO may interact in the regulation of the vascular tone. It was in fact observed that NO was able to displace bound endothelin-1 from its receptors 41 and to inhibit the synthesis of endothelin-1 in the endothelial cells. 42,43 The synthesis of endothelin-1 could be, therefore, depressed in the endothelial cells of cavernous sinuses where a strong NOS activity was to be found. It has also been shown that endothelin-1 may inhibit the production of NO in VSMC by suppressing the induction of i-NOS mRNA. 44 Thus the intense synthesis of NO in the endothelial cells of the cavernous sinuses could reduce the influence of endothelin-1 on VSMC, facilitating the expression of i-NOS. Further studies are needed to verify the hypotheses expressed in this work and to elucidate mechanisms by which endothelial cells of the cavernous sinuses, in normal subjects and in patients with vasomotor rhinitis, express more intense NOS activity compared with endothelial cells of the other vessels of the NRM.


This study, by showing a strong NOS activity in VSMC of the cavernous sinuses, accounts for the vascular disorders found in vasomotor rhinitis. Strong liberation of neuropeptides from sensory fibers by means of nerve axon responses is considered a probable factor of NOS induction in VSMC of the cavernous sinuses. Moreover, it is speculated that interactions between NO and endothelin-1 could make VSMC of the cavernous sinuses more susceptible to inducing NOS factors.