INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bradykinin (BK) and lysylbradykinin (LBK) are potent proinflammatory peptides that have been implicated as potential mediators of human allergic airway diseases (1). Increased levels of kinins have been detected in airways secretions during experimentally induced, and naturally occurring, allergic inflammation in both the upper (2) and lower airways (5, 6). Moreover, application of BK to the human airway mucosa can induce relevant symptomatic responses (7), and a BK receptor antagonist has been shown to improve pulmonary function in asthmatic subjects (13).

Interestingly, the ability of BK to induce airway symptoms varies depending upon the presence or absence of active allergic airway inflammation. Thus, while inhalation of BK causes dose-dependent bronchoconstriction, together with retrosternal discomfort and cough, in subjects with asthma, normal subjects show no changes in pulmonary function, even when high doses are inhaled (12). Airway reactivity of asthmatic subjects to BK correlates with the degree of eosinophilic inflammation (14), and can be markedly enhanced by exacerbating allergic airway inflammation using allergen challenge (15). Similarly, using a model of localized, unilateral nasal challenge with BK, we have recently shown that subjects with active allergic inflammation show nasal hyperreactivity to BK, and that this hyperreactivity is mediated, at least in part, by enhanced neural responses (16). The present studies were undertaken to further examine two specific aspects of the potential mechanisms underlying the hyperresponsiveness of inflamed airways to kinins.

First, to more fully define the role of neural mechanisms in the nasal hyperresponsiveness to BK, we examined the effects of repeated intranasal challenges with BK on various aspects of the response to challenge of subjects with perennial allergic rhinitis. The hypothesis we tested was that neural responses would show desensitization. To further confirm that selected response parameters were neurally mediated only in the setting of active allergic disease, studies were also performed in normal subjects.

The second aspect of the study was to determine if hyperreactivity of inflamed airways to kinins could be mediated via induction of human B₁-kinin receptors. Two subtypes of kinin receptors were originally proposed based upon data in animal tissues (17). In this initial definition, the B₁-kinin receptor was characterized by the fact that the carboxypeptidase metabolites of kinins, des-Arg⁹-BK and des-Arg¹⁰-lysylbradykinin (desLBK), were more potent than the parent peptides. On the B₂ receptor, BK and LBK were equally active, but the carboxypeptidase metabolites were inactive. The existence of two subtypes of kinin receptors was confirmed by the recent cloning of the genes encoding both human B₂ and B₁-receptors (18, 19). Although B₂-kinin receptors are constitutively expressed on many cell types, B₁ receptors are rarely expressed on normal tissues but can be induced in response to pathophysiological stimuli. In animals, for example, de novo expression of B₁ receptors in vivo occurs upon exposure to proinflammatory stimuli, such as cytokines, raising the implication that this receptor may play a role in the actions of kinins in chronic inflammatory conditions (20). Further support for this concept comes from studies showing that B₁-receptor antagonists have antinociceptive effects in rodent models of persistent hyperalgesia (21, 22).

It is known that a carboxypeptidase capable of converting BK and LBK to des-Arg⁹-BK and desLBK is present during allergic inflammation of the airways (23). Moreover, once the human B₁ receptor is expressed, functional responses to ligand stimulation can continue almost indefinitely, because the receptor does not undergo ligand-induced desensitization of functional responses (24). To date, however, it has been implied that the actions of kinins in the human airways are mediated via B₂-kinin receptors, because BK is a potent bronchoconstrictor in the lower airways and can induce symptoms of rhinitis when applied to the upper airways, whereas the putative B₁-receptor agonist, des-Arg⁹-BK, is inactive in both settings (25, 26). This assertion has recently been cast into doubt, however, because it has been shown that, in contrast to the receptor from several animal species, des-Arg⁹-BK is an ineffective ligand at the human receptor, and it is clear that desLBK is the natural ligand (19, 24). To reexamine the issue of the role of B₁ receptors in kinin actions in inflamed airways, therefore, we performed double-blind crossover studies comparing the effects of BK and desLBK, in the lower airways of asthmatics, and in the upper airways of subjects with symptomatic allergic rhinitis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Materials

The following materials were purchased: BK and desLBK (Bachem, Torrance, CA); normal saline solution (Abbott Labs, North Chicago, IL); lactated Ringer's solution (Kendall McGaw Laboratories, Irvine, CA); white filter cards (no. 190005; Shandon, Pittsburgh, PA); sheep antibody to human albumin and horseradish peroxidase-conjugated sheep antibody to human lysozyme (The Binding Site, San Diego, CA); rabbit antibody to human lysozyme and horseradish peroxidase- conjugated rabbit antibody to human albumin (Dako Corp., Carpinteria, CA); 96-well microtiter plates (Dynatech, Chantilly, VA); o-phenylenediamine dihydrochloride (OPD) and human serum albumin (Sigma Chemical Co., St. Louis, MO); and human lysozyme (Calbiochem, San Diego, CA). Stock solutions (10 mg/ml) of BK and desLBK for provocation experiments were prepared in lactated Ringer's solution containing an additional 1.5 g/L of sodium bicarbonate. These solutions, which had a pH of 6.5, were sterile filtered and stored in aliquots at 80° C. Further dilutions were made immediately before use in appropriate diluent solutions.

Subjects

For nasal tachyphylaxis studies, 10 subjects (six female) with perennial allergic rhinitis, defined as the presence of symptoms for at least 10 mo of the year as well as the presence of positive epicutaneous skin tests to one or more perennial allergens, were recruited. In addition, eight normal control subjects (four female), with no history of nasal symptoms and negative skin test responses to a battery of common aeroallergens were also recruited. None of the subjects had used any form of corticosteroids within 1 mo of challenge and no subject had taken astemizole within the previous 3 mo. Other antihistamines were discontinued for at least 5 d prior to challenge and decongestants were discontinued at least 24 h before provocation.

To evaluate any potential role of B₁-kinin receptors in the lower airways, six subjects (four female) with a history of asthma were recruited. All subjects had a baseline FEV₁ > 70% of predicted and had previously been shown to respond to BK bronchoprovocation with a provocative concentration causing a 20% reduction in FEV₁ (PC₂₀) of < 3 mg/ml. No subject was taking any medications at the time of challenge. For studies of the relative role of B₁ and B₂ receptors in the upper airways, eight subjects (six female) with perennial allergic rhinitis, as previously defined and with the same restrictions on medications, were recruited.

All study protocols were approved by the institutional review board of the Johns Hopkins Bayview Medical Center, and all subjects gave informed consent prior to study participation.

Study Protocols

Tachyphylaxis studies were performed in an open manner because emphasis was on the evaluation of objective indices of the nasal response. The nasal challenge protocol used was similar to that previously described (16). In brief, nasal challenges were performed, and nasal secretions collected, using 8-mm paper discs. Each disc used to collect secretions was first sealed in a 1.5-ml microtube and preweighed on an analytical balance (Mettler Instruments, Highstown, NJ) before challenge. At the beginning of the protocol, each subject received three 10-ml nasal lavages (5 ml per nostril) with saline to clear the nose of residual mucus. Five minutes later, a challenge disc containing 50 µl of saline was placed on the anterior portion of the right nasal septal mucosa. The challenge disc was removed after 1 min. Thirty seconds later, a preweighed collection disc was placed at the same spot on the challenge (ipsilateral) side and an additional disc was placed at the corresponding location in the opposite (contralateral) nostril. Both discs were removed after 30 s and resealed in their original microtube and reweighed. These discs reflect secretions recovered 2 min after challenge. After a 3-min pause, secretions were collected again in the same manner as before. This collection reflects secretions recovered 5 min after challenge. Subjects rested 4.5 min before the next challenge. Four challenges, each with a disc containing 10 µg of BK, were applied at 10-min intervals. Secretions were collected at 2 and 5 min after each challenge in a manner identical to the saline challenge. For each collection disc, precollection weights were subtracted from postcollection weights to determine final secretion weights. Proteins were eluted from discs by placing each disc in phosphate-buffered saline (PBS). For each challenge, eluates collected at 2 and 5 min were pooled and frozen. Sneezes were counted after each challenge.

A randomized, double-blind crossover protocol was used to compare the effects of B₁- and B₂-kinin receptor agonists in the airways. For each aspect of the study, challenges with the two agonists were separated by 3 to 7 d. For studies in the lower airways, a previously described protocol (15) was used. At each visit, baseline FEV₁ was assessed with the highest of three consecutive measurements being recorded. After this, 1 ml of sterile PBS was placed in a DeVilbiss Model 646 nebulizer (Devilbiss Co., Somerset, PA). Subjects then activated a Rosenthal-French model D-2A dosimeter (Lab. of Applied Immunology, Baltimore, MD) connected to compressed air at a pressure of 20 lb/ in², thereby allowing the dose to be synchronized with the subject's inspiration and delivered over 0.6 s. Subjects were instructed to take five slow breaths from functional residual volume to total lung capacity, holding each breath for 3 s. After a 3-min period, pulmonary function testing was repeated with the highest FEV₁ value serving as the baseline for the upcoming provocation. Increasing doses of BK or desLBK, administered in half-log increments beginning at a concentration of 0.003 mg/ml, were given until FEV₁ values were reduced from baseline by 20% or a concentration of 10 mg/ml was reached. Number of coughs was also noted at each dose. Subjects were permitted to leave the laboratory when FEV₁ returned to within 10% of the initial value.

For studies in the upper airways, nasal challenges were performed essentially as described previously except that, instead of four repeated challenges with 10 µg of BK, four challenges with discs containing 1, 3, 10, and 30 µg of BK or desLBK were applied at 10-min intervals and only ipsilateral collection discs were used. Subjects were asked, after each challenge, to note subjective symptom scores for rhinorrhea, nasal congestion, and irritation for each nostril using a four-point scale (zero = no symptoms, one = mild, two = moderate, three = severe). Values from both nostrils were added. To control for the possibility that symptom scores may reflect a mechanical hyperreactivity of the nasal mucosa to repeated application of discs in subjects with perennial allergic rhinitis, a third challenge arm was included in which, in a single-blinded fashion, repeated saline challenges were administered.

Protein Measurements

At the end of each nasal challenge experiment, 400 µl of PBS was added to each tube containing a collection disc. Tubes were then incubated for 30 min at 4° C to elute proteins from the discs. For each challenge, eluates from discs used to collect secretions at 2 min and at 5 min after challenge were pooled and used for analysis of proteins.

Lysozyme, an index of serous glandular secretion, was measured by ELISA. Rabbit antibody to human lysozyme was coated on 96-well microtiter plates at a dilution of 1/400 and placed at 4° C overnight. Plates were then washed and 1% sheep serum was added to each well to block nonspecific binding sites. After washing, lysozyme standard or samples at various dilutions were added to the plate and incubated at 37° C for 90 min. The plates were washed again and a 1/1,200 dilution of sheep anti-human lysozyme antibody conjugated to horseradish peroxidase was added to each well and incubated at 37° C for 90 min. The plates were washed and then developed using 0.4 mg/ml OPD as the substrate, and the absorbance was read at 490 nm. The standard curve for the ELISA ranged from 1 to 128 ng/ml.

Albumin was also measured as an index of increased vascular permeability using a similar ELISA protocol, sensitive to 1 ng/ml of albumin, as previously described (16).

Statistical Analyses

Nonparametric methods were used for data analysis (27). For the tachyphylaxis study, values for each parameter after saline challenge were subtracted from values for each of the four BK challenges. Values from the first BK challenge were considered to represent a "baseline" against which to compare subsequent challenges to evaluate if tachyphylaxis occurred. Data from the four challenges were initally compared by Friedman analysis of variance (ANOVA) to determine if significant differences were observed. Subsequent post hoc analysis used Wilcoxon matched pairs signed ranks test to compare values from each challenge to those from the first baseline challenge. For the studies on receptor subtype usage, the bronchoconstrictor capacities of BK and desLBK were assessed by their ability to reduce FEV₁, expressed as PC₂₀ calculated by linear interpolation of the dose-response curve. In the upper airways, the ability of BK or desLBK to induce significant responses above saline was first evaluated using Friedman ANOVA, with subsequent post hoc comparison for each challenge dose using Wilcoxon matched pairs signed ranks test. To compare responses between BK and desLBK, data were expressed as the sum of the net increases above saline for each dose and were analyzed by Wilcoxon matched pairs signed ranks test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tachyphylaxis Studies

To evaluate tachyphylaxis of neural responses, we initially studied subjects with perennial allergic rhinitis. In these subjects, sneezing showed significant tachyphlyaxis upon repeated challenges with BK (p < 0.01 by Friedman ANOVA). Sneezing was only significantly increased (p < 0.04) above diluent for the first BK challenge, when a median of 1.5 sneezes (25th percentile = 0; 75th percentile = 5) was recorded. For each subsequent challenge, sneezing was significantly reduced compared with the first challenge (p < 0.05), with the median sneeze count in each case being zero.

In subjects with perennial allergic rhinitis, significant increases above saline control were seen in both ipsilateral and contralateral secretion weights (p < 0.01 in each case) after the first BK challenge (Figure 1). When analyzed as increases above saline compared with the first BK challenge, tachyphylaxis was observed upon subsequent challenges for both ipsilateral and contralateral secretion weights (p < 0.01 by ANOVA in each case). Subsequent post hoc analyses confirmed that each of challenges 2 through 4 produced responses for both ipsilateral and contralateral secretion weights that were significantly less than those for the first challenge. Indeed, no significant increases above saline were observed in contralateral secretion weights for any subsequent challenge, whereas for ipsilateral secretion weights, only BK challenge 3 induced a significant increase above saline challenge (p < 0.05). The ipsilateral secretion weight for challenge 3 was not significantly higher, however, than the responses for challenges 2 and 4. 

View larger version (22K): [in this window] [in a new window]   Figure 1.   Effects of repeated unilateral challenges of subjects with perennial allergic rhinitis with BK on ipsilateral (Ipsi) and contralateral (Contra) secretion weights. Data for each challenge are expressed as increase above saline challenge. Median values for secretion weights after saline challenge were 15.1 mg (Ipsi) and 21 mg (Contra). Negative values indicate responses that were lower than that for the saline control challenge. Columns represent median values; numbers in parentheses are 25th and 75th percentile values. Asterisks indicate that values are significantly lower than those for challenge 1.

The first BK challenge also induced significantly increased concentrations of the serous glandular marker, lysozyme, in both ipsilateral and contralateral secretions (p < 0.01 in each case) compared with saline challenge (Figure 2). Marked tachyphylaxis was seen, however, for induction of ipsilateral (p < 0.01 in each case) or contralateral (p < 0.02 in each case) lysozyme upon subsequent challenges when compared with the initial challenge. Moreover, none of challenges 2 through 4 caused a significant increase above saline challenge for ipsilateral or contralateral lysozyme.

View larger version (18K): [in this window] [in a new window]   Figure 2.   Effects of repeated unilateral challenges of subjects with perennial allergic rhinitis with BK on ipsilateral (Ipsi) and contralateral (Contra) levels of lysozyme. Data for each challenge are expressed as increase above saline challenge. Median values for lysozyme concentrations after saline challenge were 4.9 µg/ml (Ipsi) and 3.3 µg/ml (Contra). Negative values indicate responses that were lower than that for the saline control challenge. Columns represent median values; numbers in parentheses are 25th and 75th percentile values. Asterisks indicate that values are significantly lower than those for challenge 1.

Because we have previously shown that, even in subjects with perennial allergic rhinitis, there is no specific BK-induced increase in concentrations of albumin in contralateral secretions (16), albumin levels were monitored only in recovered ipsilateral secretions. As expected, the initial BK challenge caused a striking and significant (p < 0.01) increase in albumin levels compared with saline challenge (Figure 3). Although each subsequent challenge also induced significant increases (p < 0.02 in each case) in albumin levels compared with saline challenge, when the increases above saline control were compared with the initial challenge, significant tachyphylaxis was observed (p < 0.005 by ANOVA). Both the second and fourth challenges were significantly reduced compared with the initial challenge (p < 0.02 in each case). This finding was surprising, because it has generally been assumed that the ability of BK to induce increased vascular permeability is due to a direct effect on receptors on the vascular endothelium. To determine if the observed reductions in albumin concentrations may represent tachyphylaxis of a neurally mediated component of vascular leak specifically present in subjects with active disease, or resulted from a more nonspecific effect, such as receptor desensitization or internalization (28), we examined the response to repeated challenge in normal subjects.

View larger version (13K): [in this window] [in a new window]   Figure 3.   Effects of repeated unilateral challenges of subjects with perennial allergic rhinitis with BK on ipsilateral levels of albumin. Data for each challenge are expressed as increase above saline challenge. Median value for albumin concentrations after saline challenge was 88 µg/ml. Negative values indicate responses that were lower than that for the saline control challenge. Columns represent median values; numbers in parentheses are 25th and 75th percentile values. Asterisks indicate that values are significantly lower than those for challenge 1.

Consistent with our prior studies, provocation of normal subjects with BK induced a more modest secretory response than that observed for subjects with perennial allergic rhinitis (Figure 4). Significant increases above saline control were observed for ipsilateral secretion weight at each of the four BK challenges (p < 0.03 in each case), and there was no difference in the magnitude of the responses at each dose (p = 0.47 by ANOVA). No significant increases above saline control were observed for contralateral secretion weights at any dose (p = 0.56 by ANOVA). Lysozyme concentrations were not measured because we have previously shown that there is no significant serous glandular secretion induced by BK challenge in normal subjects (16).

View larger version (13K): [in this window] [in a new window]   Figure 4.   Effects of repeated unilateral challenges of normal subjects with BK on ipsilateral (Ipsi) and contralateral (Contra) secretion weights. Data for each challenge are expressed as increase above saline challenge. Median values for secretion weights after saline challenge were 6.7 mg (Ipsi) and 9.4 mg (Contra). Negative values indicate responses that were lower than that for the saline control challenge. Columns represent median values; numbers in parentheses are 25th and 75th percentile values. Asterisks indicate that values are significantly lower than those for challenge 1.

Significant (p < 0.005 by ANOVA) increases in albumin concentrations were observed in ipsilateral secretions in response to BK (Figure 5). Albumin levels were elevated above saline control levels for each BK challenge (p < 0.02 in each case), but the response above saline control at the initial BK challenge was significantly lower than for the comparable challenge in subjects with perennial allergic rhinitis (p < 0.02 by Mann-Whitney U test). In contrast to data for the subjects with perennial allergic rhinitis, there was no tachyphylaxis upon repeated challenge, as evidenced by the lack of significant difference between the four responses (p > 0.1 by ANOVA).

View larger version (13K): [in this window] [in a new window]   Figure 5.   Effects of repeated unilateral challenges of normal subjects with BK on ipsilateral levels of albumin. Data for each challenge are expressed as increase above saline challenge. Median value for albumin concentrations after saline challenge was 15 µg/ ml (Ipsi). Negative values indicate responses that were lower than that for the saline control challenge. Columns represent median values; numbers in parentheses are 25th and 75th percentile values.

Agonist Studies

The responses of the asthmatic subjects to bronchoprovocation with BK and desLBK are summarized in Table 1. As expected, BK challenge induced bronchoconstriction in all six subjects, with the PC₂₀ being less than 2,000 µg/ml in all cases. By contrast, no bronchoconstriction was observed using desLBK. Not only could we not obtain a PC₂₀ for desLBK in any subject with the highest dose used but, even at this highest dose (10 mg/ml), the measured FEV₁ (median = 3.1 L) was not different from that after saline challenge (median = 3.08 L; p = 0.67). Cough responses were also prominent with BK but not with desLBK. Any analysis of cough counts needs to take into account the fact that subjects always received fewer doses of BK than desLBK. In Table 1, data are presented as total number of coughs divided by number of doses of each agonist received. An alternative approach is to express data as total coughs for the same number of doses of the two agonists, defined as the number of doses of BK received. When analyzed in this fashion, marked differences were again observed with BK causing a median of 15 coughs (25th percentile = 8; 75th percentile = 26), whereas desLBK caused a median of 0 coughs (25th percentile = 0; 75th percentile = 3).

In the upper airways, subjects recorded increases in subjective symptoms regardless of the challenge stimulus. Total symptom scores from four repeated saline challenges (median = 10) were not different (p = 0.24 by Friedman ANOVA) from the scores recorded from the challenges with either desLBK (median = 10) or BK (median = 13), indicating nonspecificity of this outcome measure. By contrast, while BK caused a significant dose-dependent increase in secretion weights compared with saline control challenge (p < 0.001 by ANOVA), desLBK did not induce a significant increase (p = 0.45) in secretion weights (Figure 6). The B₁-receptor agonist, desLBK, also failed to induce serous glandular secretion (Figure 7), since lysozyme was not significantly increased above saline control (p = 0.87), whereas the B₂-receptor agonist, BK, induced a dose-dependent increase in lysozyme secretion (p < 0.001). There was also no evidence of a role for B₁-receptor stimulation in increasing permeability of the nasal vasculature (Figure 8). In contrast to BK, which caused striking increases in concentrations of albumin in secretions (p < 0.001), desLBK caused no significant increases above saline control (p = 0.57). Repeated saline challenges had no effect on any of these objective indices in symptomatic allergic subjects, and desLBK had no effects on any parameter in normal subjects (not shown).

View larger version (26K): [in this window] [in a new window]   Figure 6.   Effects of unilateral challenges of subjects with perennial allergic rhinitis with BK or desLBK on ipsilateral secretion weights. Data are expressed as increase above saline challenge. Median values for secretion weights after saline challenge were 9 mg (BK) and 16 mg (desLBK). Negative values indicate responses that were lower than that for the saline control challenge. Columns represent median values; numbers in parentheses are 25th and 75th percentile values.

View larger version (24K): [in this window] [in a new window]   Figure 7.   Effects of challenges of subjects with perennial allergic rhinitis with BK or desLBK on ipsilateral lysozyme concentrations. Data are expressed as increase above saline challenge. Median values for lysozyme concentrations after saline challenge were 6.8 µg/ml (BK) and 8.4 µg/ml (desLBK). Negative values indicate responses that were lower than that for the saline control challenge. Columns represent median values; numbers in parentheses are 25th and 75th percentile values.

View larger version (21K): [in this window] [in a new window]   Figure 8.   Effects of challenges of subjects with perennial allergic rhinitis with BK or desLBK on ipsilateral albumin concentrations. Data are expressed as increase above saline challenge. Median values for albumin levels after saline challenge were 34 µg/ml (BK) and 68 µg/ml (desLBK). Negative values indicate responses that were lower than that for the saline control challenge. Columns represent median values; numbers in parentheses are 25th and 75th percentile values.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have used a model of localized nasal challenge to confirm and extend our earlier observations that nasal hyperreactivity to BK in the setting of active allergic disease is mediated, at least in part, by increases in neural responsiveness (16). Our current data clearly demonstrate that repeated challenge with BK leads to a desensitization of neurally mediated components of the response. Thus, sneezing, a neural reflex, was observed on the first challenge with but not upon subsequent repeated challenges with the same dose. The marked increases in secretion weights and in serous glandular secretion noted in subjects with perennial allergic rhinitis also occur secondary to neural stimulation. These responses have previously been shown to be mediated via parasympathetic cholinergic reflexes and can be inhibited by atropine pretreatment (16). Repeated BK challenges also led to tachyphylaxis of these secretory responses, with both ipsilateral and contralateral secretion weights and ipsilateral and contralateral levels of the serous glandular marker, lysozyme, being significantly reduced after the first challenge.

The reduction in concentrations of albumin, a marker of vascular permeability, that was observed upon repeated challenge of subjects with perennial allergic rhinitis was surprising. Kinins have been known for many years to be among the most potent inducers of increased vascular permeability, but this is generally attributed to direct interaction of kinins with receptors on vascular endothelium (1). Indeed, a previous study had failed to show tachyphylaxis of BK-induced increased vascular permeability in normal and asymptomatic atopic subjects upon repeated spray challenges (26). The tachyphylaxis seen in perennial allergic subjects was not due simply to the use of a localized challenge model because we did not see reductions in albumin levels when normal subjects were repeatedly challenged using the same method (Figure 5). This selective tachyphylaxis in subjects with perennial allergic rhinitis raised the possibility that a neural component may contribute to the vascular response in these subjects. This was supported by the observation that albumin concentrations were significantly higher after BK challenge in allergic subjects compared with normal control subjects, although it must be noted that such a difference was not observed in an earlier study (16). The reasons for this difference are unknown but may represent variabililty among the subject populations. The data from the current study, however, are consistent with the concept that a direct effect of BK on the vasculature is supplemented, in subjects with perennial allergic rhinitis, by a neurally mediated pathway that also induces permeability. Upon desensitization of this component, responses then resemble those in normal subjects. Because BK does not increase vascular permeability in the nostril contralateral to the unilateral challenge (16), any neural contribution to increased vascular permeability at the challenge site presumably reflects the local release of neuropeptides rather than a central reflex pathway. There is precedent for allergic inflammation to increase neuropeptide expression in afferent sensory nerves in guinea pigs (29), and it has been shown that nasal challenge of allergic humans with BK can induce the release of the neuropeptide, substance P (30). Furthermore, although an earlier study using nasal challenge with a low dose of capsaicin, a potent stimulus of sensory nerves, did not find increased vascular permeability in allergic subjects (31), more recent data, using higher doses, demonstrate that capsaicin induces increased vascular permeability in human subjects with perennial allergic rhinitis, but not in subjects with nonallergic rhinitis or in normal control subjects (32). Moreover, this effect of capsaicin was inhibited by blocking sensory nerve activation using the topical anesthetic, lidocaine (33).

The mechanisms underlying the ability of kinins to induce neural reflexes in the setting of active allergic airway disease, but not in healthy airways, remains to be determined. In this study we tested the hypothesis that inflammation induces neural hyperresponsiveness by inducing B₁-kinin receptors. This receptor subtype is rarely expressed on normal tissues but studies with human cells (19), as well as in intact animals (20), have shown that it can be induced upon exposure to proinflammatory stimuli. Although it previously has been shown that the putative B₁-receptor agonist, des-Arg⁹-BK, did not exert effects in the human airways (26), recent studies with the cloned human B₁ receptor demonstrated that des-Arg⁹-BK is an extremely weak agonist, and that desLBK, which is over 100-fold more potent, is the natural ligand (24). Moreover, the earlier studies using des-Arg⁹-BK in the upper airways were performed in subjects without active allergic airway disease (26).

We chose to examine the effects of desLBK in asthmatic subjects who had already demonstrated good airways reactivity to BK because of prior evidence indicating that the ability of BK to induce bronchoconstriction is related to the degree of airway inflammation (14, 15). We reasoned that B₁-receptor induction was most likely to occur in such actively inflamed subjects. Despite this attempt to favor conditions for B₁-receptor induction, however, no bronchoconstriction was observed in any of the asthmatic subjects in response to desLBK, whereas marked responses to BK were confirmed. Although these experiments showed that B₁ receptors do not play a role in the bronchoconstrictor effects of kinins in inflamed airways, they did not provide information on other aspects of the airway response, such as glandular secretion and increases in vascular permeability. Our studies in the upper airways of subjects with perennial allergic rhinitis, however, allowed us to directly assess the role of B₁-receptors in these aspects of airway responsiveness to kinins.

Increases in nasal symptom scores were noted upon nasal challenge with either BK or desLBK. Given that challenge stimuli are applied to a small (8-mm), localized area of the nasal mucosa, it is not surprising that symptom scores were relatively low. It was of interest to note, however, that similar increases were observed even with repeated saline challenges, suggesting that neither B₁- not B₂-kinin receptor stimulation increased symptoms in a specific manner in this model. This raises the possibility that symptom scores recorded when subjects with perennial allergic rhinitis are challenged in this model mainly reflect responses to local mechanical irritation induced by repeated application of challenge, and collection, discs. We have previously shown that symptoms are markedly lower in normal subjects after BK challenge compared with subjects with active rhinitis (16), suggesting that the hyperresponsiveness to mechanical irritation noted in the current study occurs only in inflamed airways. This is consistent with data showing that allergic inflammation induces mechanical hyperreactivity of afferent neurons in guinea pig airways (34).

In contrast to data for subjective symptom scores, there were clear differences between the responses to B₁- and B₂- receptor stimulation when objective indices of serous glandular secretion and vascular permeability were examined. Challenges with desLBK did not increase secretion weights nor concentrations of lysozyme or albumin in recovered secretions. By contrast, BK challenge led to striking increases in each of these parameters even at the lowest dose used. Given our data that repeated challenges with BK can lead to tachyphylaxis of responses, it is likely that even higher responses to the top dose of BK would be observed if preceding challenges with lower doses had been omitted.

In summary, therefore, our data suggest that subjects with perennial allergic rhinitis show enhanced neural responsiveness. This is manifested both as increased responsiveness to mechanical stimulation, leading to increased symptom scores, and as hyperreactivity to challenge with BK. Whereas central parasympathetic reflexes are involved in increases in glandular secretion, we provide the first evidence implying that a local "axon reflex" may contribute to the ability of BK to increase local vascular permeability in inflamed subjects. Both these types of neural responses show desensitization upon repeated challenges. The hyperresponsiveness to kinins that is seen in human airways, in subjects with active allergic disease, cannot be attributed to increased expression of functional B₁-kinin receptors, because the selective B₁-receptor agonist, desLBK, had little or no effect in either the upper or lower airways of such subjects. The mechanisms by which neural hyperresponsiveness to kinins is induced in subjects with active allergic disease, therefore, remain to be determined.