Hyperresponsiveness in the human nasal airway: new targets for the treatment of allergic airway disease.

Allergic rhinitis is a condition which affects over 15% of the population in the United Kingdom. The pathological process involves two stages: nasal inflammation, and the development of nasal airway hyperresponsiveness (AHR) to allergen and a number of other stimuli. This results in the amplification of any subsequent allergic reaction, contributing to the chronic allergic state. A number of different hypotheses have been proposed to explain the underlying mechanism of AHR, including a role for eosinophil-derived proteins, free radicals and neuropeptides. While there may be a number of independent pathways which can result in AHR, evidence obtained from both animal models and in vivo experiments in humans indicate that some mediators may interact with one another, resulting in AHR. Further research into these interactions may open new avenues for the pharmacological treatment of chronic allergic rhinitis, and possibly other allergic airway diseases.


Introduction
Alle rgic rhinitis and asthma are tw o of the most common immunological disorde rs produc ing chronic disease in man. For instance, alle rgic rhinitis affec ts over 15% of the gene ral population. 1 In both dise ases, a loc al allergic reaction results in an inf lammatory response, causing w heezing and muc us sec retion in asthma, and nasal congestion, rhinorrhoea ('runny nose '), pruritis (itching) of the nose and sne ezing in alle rgic rhinitis. Continue d ex posure to allergen induc es a state of chronic allergic inflammation. Contributing to this is the de ve lopme nt of airw ay hyperresp onsive ness (AHR), w hich re sults in the amplification of any subsequent allergic reaction.
While a variety of pharmac ological age nts can be used to treat the initial inflammatory response, only steroids are effe ctive in preventing AHR. The mechanism by w hich AHR oc curs re mains uncle ar. If the me chanism of AHR can be de scribed, then this may provide new pharmac ological targets for the treatme nt of chronic allergic airw ay disease. While allergic rhinitis is not life-threatening, the same is not true for chronic asthma, and the problem is compounded by the adve rse effe cts associated w ith long-te rm steroid use.
Unfortunate ly, animal models and in vitro ex perime nts are of limited value, since the y are not repre sentative of the situation in man. 2 ,3 How ever, the human nasal airw ay is re asonably acc essible for in v ivo rese arch into the de ve lopme nt of nasal AHR, and may also yie ld some insight into the me chanism of AHR in asthma. We re view he re c urre nt hypotheses for the de ve lopme nt of AHR in the human nose, and indicate potential new pharmacologic al targe ts for the tre atment of alle rgic rhinitis and possibly other alle rgic airw ay dise ase s.

Pathogenesis of allergic rhinitis
There are tw o distinct forms of allergic rhinitis, depending on the allergen to w hich the subje ct is se nsitive. In pere nnial allergic rhinitis (PAR), the subject is ex posed to allergen throughout the ye ar, w hile in subjec ts w ith se asonal allergic rhinitis (SAR), alle rgen ex posure is limite d to c ertain times during the year. Subjec ts w ith PAR are sensitive to alle rge ns including those from the house -dust mite (De rm a toph a g o ide s pte ro n y s s inu s ), other mites, animal danders and c ertain foods. In c ontrast, the allerge ns involved in SAR are tre e and grass pollens, moulds and ce rtain fungal sp ores.

Nasal inflammation
Whe n particles of allergen are inhaled into the nasal cavity, the y cross the nasal muc osa and are ex posed to the immune syste m; in susc eptible individuals, this results in sensitisation, a process w he re immunoglobulin E (IgE) is synthe sise d and bec omes fix e d to ce lls, such as mast ce lls. If the se c ells subse que ntly encounter the same allergen, c rosslinking oc curs betw e en IgE molec ule s on the c ell surface , c ausing the release of inflammatory mediators including histamine, kinins and eic osanoids, such as PGD 2 , 4 as show n in Table 1. The se mediators are resp onsible for the symptoms of alle rgic rhinitis. Nasal blockage results from a combination of incre ased blood flow to the nose (causing sw e lling of the nasal tissue), plasma ex travasation, and oedema. Rhinorrhoe a is a conse que nce of the stimulation of nasal sec retory glands, w hich produc e a mix ture of w atery or se rous se cretion and visc ous mucus se cretion, mix e d together w ith ex travasated plasma. Stimulation of se nsory nerves in the nasal cavity causes snee zing and pruritis.
One me thod use d to study the inflammatory response in the human nose is nasal lavage, w here the nasal cavity is rinsed w ith a solution (such as saline ), and the lavage fluid subsequently colle cted and analysed. Using this te chnique , one c an measure a number of inflammatory me diators using ap propriate assays. Furthermore, by collecting and c ounting c ells present in the lavage fluid, it is possible to correlate the rele ase of inflammatory me diators w ith the different cells observe d.
The response of subje cts w ith allergic rhinitis to intrana sal challenge can often be divided into an imme diate phase, oc curring during the first 2 h follow ing ex posure to antigen, and, in about 30-40% of subje cts , a late phase occurring 6 -12 h later. The late phase is associated w ith an infiltration of a number of differe nt inflammatory cells (including eosinophils, ne utrophils and T-lymphocyte s) to the site of inflammation, w here the y re le ase a variety of me diators including leukotriene s, c ationic prote ins such as eosinophil c ationic p rote in (ECP), and cytokines (Table 1).
Intere stingly, the re may be differe nce s in the prec ise mediators involve d in PAR and SAR. While histamine is strongly associated w ith the early p hase of SAR, histamine H 1 antagonists are not very e ffec tive in reducing the early allergic re sponse of PAR.
Inste ad, the kinins appe ar to be the major mediator causing symptoms in PAR. 5 There is also e vide nce that the allergen (house dust mite ) in PAR c an ge nerate kinin indepe nde nt of the inflammatory response, either by the dire ct ac tivation of kallikrein, the enzyme w hic h ge nerates kinins, 6 or by posse ssing such biochemic al activity itself. 7

Nasal airway hyperresponsiveness
Nasal hyperresponsivene ss is a hallmark of allergic rhinitis . 8 ,9 Subje cts w ith alle rgic rhinitis show an inc re ased response to nasal challenge w ith a variety of stimuli, including histamine and bradykinin, both of w hich are re le ased follow ing allergen challe nge. The nasal airw ay e ffec tively bec omes more se nsitive to alle rgen, c ontributing to the chronic alle rgic state. AHR is usually assoc iate d w ith the late p hase reaction, but c an continue w ell be yond this stage. In fac t, it is induc ed irresp ective of w hethe r a late phase of inflammation occ urs. 10 One important c onside ration is the typ e of response involved in AHR. In asthma, AHR is often defined as an inc rease in the contrac tile resp onse of the airw ay smooth muscle. How e ver, the re is little smooth muscle prese nt in the human nasal airw ay. Nasal blockage results from inc reased blood flow to the nasal mucosa, and not by an action on smooth muscle. This highlights the difficulty of ex trapolating data from the nasal airw ay to the low er airw ays, and vic e ve rsa. AHR is also assoc iated w ith increase d mucus production and oe de ma follow ing allergen challe nge, in both the upper and low e r airw ays. Therefore, AHR w ould appe ar to result from a dire ct or indirect pote ntiation of the ove rall rece ptor activation, and not simply from an ac tion on a particular ce ll type alone . With this in mind, there are a number of pote ntial mechanisms by w hich AHR might oc cur ( Fig. 1).
(1) Th e in fla m m a to ry c a s c a de : Initial ex posure to alle rgen might trigge r a casc ade reaction, c ausing inc reased mediator re le ase and, therefore, gre ate r rec eptor activation. For instanc e, subjects w ith SAR ex perie nce inc re asingly se vere symptoms as the polle n se ason continue s. How ever, a number of inflammatory me diators can induce AHR in the absence of any othe r mediator. Furthermore, the development of AHR can oc cur independe ntly of a chronic inflammatory re sponse . 10 It is likely, therefore , that the AHR potentiate s the effe ct of the inflammatory c ascade se en in chronic alle rgic rhinitis, rathe r than being caused by it.
(2) In c re a s e d ex po s u re o f re ce pto rs o n th e n a s a l m u co s a to the s tim u lu s : A common histologic al finding in chronic airw ay alle rgy is the damage and shedding of e pithelial c ells lining the airw ay. The airw ay e pithe lium can act as a physical barrier betw ee n the airw ay lumen and the underlying mucosa. The re fore, removal of these cells w ould increase the ex posure of the rece ptors on the mucosa to any stimulus present. For ex ample , in the low er airw ays, removal of the airw ay ep ithe lium inc re ases the response of smooth muscle in both animal models and human bronchi ex vivo . 11 While there may be a link betw e en e pithe lial she dding and the presenc e of AHR in alle rgic individuals, 12 mediators w hich induc e AHR on the ir ow n (e .g. plate letactivating factor) do not c ause e pithelial shedding in the time take n for AHR to de ve lop, though the doses required are tox ic to airw ay e pithe lial c ells. 13 Alte rnatively, the loss of airw ay e pithe lial c ell func tion, but not ne ce ssarily barrier func tion, may be sufficie nt to induc e AHR. 1 4 In the human nasal airw ay, AHR to bradykinin is unlikely to be due to a re duction in epithe lial barrie r function, 15 and the re is no evidence for increase d ep ithe lial perme ability in subjec ts w ith alle rgic rhinitis. 1 6 Me diators such as PAF may also damage the mucoc iliary clearance syste m linked to the airw ay e pithe lium, c ontributing to the AHR. 13 (3) Mo du la tin g a g e n ts : The ability of ace tylcholine to cause vasodilatation in the vascular system is de pendent upon the re lease of nitric ox ide from the endothe lium (and possibly other me diators, colle ctive ly know n as endothe liumderive d relax ing fac tors). It has been proposed that airw ay e pithelial c ells may gene rate me diators, such as arachidonic ac id metabolites, nitric ox ide and the putative e pithelium-de rive d re lax ing fac tor (EpDRF) (also know n as EDHF (e pithelium-de rive d hyp erpolarsing fac tor)) w hich modulate the re sponsivene ss of the airw ays. How ever, the evide nce for such me diator release from airw ay epithelial c ells is equivocal. 1 1,17 -1 9 The epithe lium may have an important role in the metabolism of various stimuli, partic ularly ade nosine , 2 0 tachykinins 1 8 and ace tylcholine . 21 Loss of e pithe lial ce ll function might reduce the metabolism of the se me diators, enhancing their ability to activate rece ptors and e ffec t a response. How ever, epithe lial damage c an oc cur w ithout the prese nce of AHR, and vice ve rsa. 22 For ex ample , cationic prote ins can cause AHR w ithout any appare nt damage to the epithe lium. 2 3 (4) Re c epto rs : AHR may result from a change in the characteristic s of rec eptors loc ate d on the nasal mucosa. For ex ample, cholinergic agonists such as me thacholine c ause more se cretion in alle rgic subjects than in non-allergic controls, and this c ould be ex plained by an inc rease in the de nsity of choline rgic rec eptors. 24 How ever, the de nsity of musc arinic re ce ptors on nasal tissue from allergic subjects may actually de crease slightly, possibly as a result of adaptation due to overstimulation, although the remaining rec eptors ex hibit an inc re ased affinity, w hich might contribute to the hyperse cretion. 25 No signific ant differe nce s have bee n found for a -or b -adre noc eptors. 26 The minor changes observe d are probably a re sult, rather than the cause, of AHR.

(5) Me ta bo lis m :
Alternatively, AHR may be a c onsequenc e of prolonge d activity of a varie ty of me diators, due to an e ffec t on their metabolism. For ex ample, neuropeptide s are degrade d by neutral e ndope ptidase (NEP), and the activity of NEP is signific antly low e r in subje cts w ith nasal AHR. 27 A sec ond protein, lac tofe rrin, may preve nt antige n-induc ed AHR in the low e r airw ays of shee p by inhibiting tryptase. 2 8 Though the pre cise role of tryptase in airw ay alle rgy is not fully know n, it c an induc e AHR in the low e r airw ays of shee p 29 and in e x v ivo tissue from human bronchi. 3 0

(6) Alte re d in tra c e llu la r s ig n a llin g :
Finally, the intrac ellular pathw ays w hich are activate d by the rec eptor/agonist complex may be altered in AHR. Of partic ular interest is phosphodieste rase IV (PDE4), an enzyme w hich catalyses the metabolism of a number of intrace llular messenge rs, such as cyclic GMP (c GMP). While PDE4 inhibitors abolish AHR in animal mode ls, 31 the data in human asthmatic subje cts are e quivocal, 3 2 and the re are no data w ith respect to the human nasal airw ay. Inte re stingly, inhibitors of nitric ox ide synthase can also c ause AHR, possibly by reducing the amount of cGMP. 33 AHR could re sult from a loss of a c ompe nsatory mechanism, involving c GMP; such a mechanism may be fre e radical de pendent. 34 In the chronic allergic state , AHR may result, in part, from a change in gene transcription w ithin c ells (e.g. the NF-kB transc ription syste m), though the timesc ale of onset of hyperre sponsivene ss make s this p roc ess unlike ly in the induction of AHR.

Mediators implicated in the development of AHR
In 1933, Sir Henry Dale described a numbe r of crite ria w hich should be used to identify pote ntial inflammatory mediators. Applying the se to the development of AHR: The process by w hich AHR occ urs re mains uncle ar, no doubt due, in part, to the complex ity of the me chanism and the diffic ulty in regulating the process under c ontrolled ex pe rimental conditions. Nonetheless, bec ause it is assoc iated w ith the influx of ce lls to the site of inflammation, a number of hypotheses have be en p ropose d involving me diators release d from the se c ells. One partic ular c ell type, the eosinophil, has been implic ate d in AHR. In both SAR and PAR, antige n challenge re sults in an inc rease of eosinophils in the nasal mucosa and the rele ase of various eosinophil-derived mediators. 35 -39 Cons equently a major role has been propose d for eosinophils and the ir proteins in this proc ess.

Eosinophil granule-derived proteins
Eosinophils contain granule s composed of four basic prote ins. The c ore of these granule s is major basic prote in (MBP), w hile the matrix surrounding the core is c omposed of eosinophil c ationic protein (ECP), eosinophil-derived ne urotox in (EDN) and eosinophil perox idase (EPO). 40 The possible role s of these prote ins in allergic airw ay disease are described in Table 2.
The levels of ECP, EPO and MBP are raise d follow ing antige n challe nge in alle rgic rhinitis, 37 -3 9,4 1 and these inc re ases often coincide w ith the presenc e of AHR. MBP and othe r c ationic prote ins, including the synthetic protein poly-L-lysine, cause AHR in the low er airw ays of rats, 23 w hich appe ars to be dependent on the ir c ationic charge. How ever, no study has ye t bee n conducte d to investigate w hether the se cationic proteins can the mselves induc e AHR in the human nasal airw ay.
The mechanism of MBP-induc ed AHR is unknow n, though it may be de pendent on causing epithelial damage , since MBP only induce s AHR in guinea pig tracheal preparations w ith an intac t ep ithe lium. 4 2 How e ver, as stated above , c ationic prote ins may induc e AHR w ithout c ausing e pithe lial damage. 2 3 Furthe rmore , although all the e osinophilic cationic prote ins are cytotox ic to the airw ay ep ithe lium, 4 3 only MBP caused AHR in a study on primate s. 4 4 In the low er airw ay of the rat, AHR induce d by MBP or poly-L-lys ine is abolishe d by both ne urokinin NK-1 4 5 and bradykinin B2 4 6 antagonists , indicating a role for substanc e P and the kinins in the de velopment of AHR. It is also possible that eosinophil cationic prote ins act on othe r inflammatory c ells to ge nerate the c onditions required for AHR. 2 2 While there is e vide nce for a relationship betw e en ECP levels in nasal lavage and nasal AHR, 47 this has not alw ays be en reproduc ed in other studie s, 48 ,4 9 so ECP is unlikely to be solely responsible for the induc tion of nasal AHR.

Platelet activating factor (PAF)
PAF is a naturally occ urring phospholipid, and is the only e ndogenous compound know n to induce AHR in both animals and man. 50 In normal, non-atopic subjects, it induce s a nasal AHR to histamine and bradykinin 5 1,5 2 and causes an inc re ased re sponse to polle n in subjec ts w ith SAR. 51 The re sulting AHR is similar, in many w ays, to that observe d in allergic rhinitis, as PAF also c ause s a signific ant nasal ne utrophilia and eosinophilia, toge the r w ith raise d ECP levels in nasal lavage fluid. 52 -54 Although PAF c an induc e AHR in the human nasal airw ay, its role in allergen-induced AHR is less de fine d. Antigen challe nge w ith grass polle n in atopic individuals w ith SAR causes the release of lyso-PAF and PAF, 41 ,55 but it is uncle ar w he ther PAF or lyso-PAF are release d in PAR, 56 One possibility is that any PAF generated is c onve rted to lyso-PAF by acetylhydrolase present in the lavage fluid. The ac tivity of this enzyme in lavage fluid is signific antly raised follow ing grass polle n challenge in sensitive subje cts. 41 ,57 PAF may ac t direc tly on the nasal muc osa, and radioligand binding studies indicate the p re sence of spe cific binding site s for PAF in human lung tissue, 5 8 but no such studies have yet bee n p erformed using tissue obtaine d from the human nose . One of the main fe atures of PAF is its ability to attract and ac tivate a variety of inflammatory c ells, including eosinophils, ne utrophils, monocyte s, mac rophage and platele ts. 5 0 PAF may also release from airw ay e pithe lial c ells a range of chemotactic fac tors for neutrophils and eosinophils, including the cytokine GM-CSF. 5 9 The subse que nt ac tivation of these ce lls w ill re le ase a range of mediators w hic h can act on othe r ce lls in the airw ay. Administration of PAF into the nasal airw ay also causes ECP re le ase, w hich c ould then c ontribute to the AHR. 5 2,53 If PAF is an important mediator of nasal AHR, one w ould ex pe ct PAF-antagonists to reduc e leukoc yte infiltration and inhibit antige n-induc ed AHR. While this e ffec t has bee n obse rved in some studie s using animal models, 6 0 -6 2 the PAF antagonists WEB 2086 4 1 and UK 74,505 63 did not alter AHR in the human nasal airw ay. How e ve r, both these antagonists ex hibit a low pote ncy for PAF re ce ptors, and may c ause only a w eak inhibitory ac tion in man.
PAF c auses mucosal dysfunction and damage, inhibiting ciliary action and inc re asing ex foliation of the epithelial lining of the airw ay both in vivo in the low er airw ays of the rabbit 13 and in v itro using ex plants from human nasal muc osa. 6 4 Intere stingly, the latter study found that PAF is itself c ytotox ic to epithe lial c ells, w ithout requiring the production of othe r cytotox ic mediators. PAF-induce d AHR in the human nasal airw ay w as almost abolishe d by pre treatme nt w ith the antiox idant vitamin E, 52 implying a me chanism w hich is de pendent on the gene ration of free radic als. This supp orts the hypothesis that PAF causes tissue damage w hich may be independe nt of the release of other me diators, perhap s via the generation of fre e radic als. The sourc e of the fre e radicals c ould be the e pithe lial ce lls, so in effec t, the epithe lium may cause its ow n destruc tion. 65 It is unclear, how eve r, w hether sufficie nt quantities of PAF are produce d in the antige n-induc ed allergic response to cause AHR, and there are probably differences in the underlying mechanism of AHR induc ed by PAF and antigen.

Leukotrienes
The leukotrienes are generated by the action of 5-lipox ygenase on arachidonic ac id. The y are release d in both the early and late phase s follow ing antigen challe nge in subjec ts w ith SAR 66 and during the early phase in PAR. 67 There are tw o classes of leukotriene s: LTB 4 and the p eptidyl-cyste inyl le ukotrie nes (LTC 4 , LTD 4 and LTE 4 ). The latte r group are synthe sise d by eosinophils (among other c ells) and cause a longlasting eosinophilic infiltration, and have be en associate d w ith AHR in the low er airw ays in rats 6 8 and in man; 69 both these ac tions appear to be depe ndent on eosinophil activation. Although inhibitors of le ukotriene synthesis re duce the nasal blockage ex perienced follow ing challenge w ith grass pollen in alle rgic subjec ts, 7 0,7 1 no studies have investigate d the effect of such drugs on AHR in the human nasal airw ay, though zileuton, a lipox ygenase inhibitor, reduced antigen-induc ed AHR in the low e r airw ays of asthmatics. 7 2 Pharmacologic al intervention can also be achie ved at the leve l of the rec eptor, and a number of leukotriene rec eptor antagonists have be en developed w hich inhibit AHR in animal models. 73 Many of these are currently under study in man, including pranlukast, w hich may reduce AHR in asthmatic s, 7 4 and monte lukast, w hich has rece ntly bee n approve d for use in the USA. The role of LTB 4 in AHR is not cle ar, but LTB 4 antagonists have be en show n to inhibit antigen-induce d AHR in the low e r airw ays of primates 75 and the guine a pig. 76 Although human eosinophils cannot produce LTB 4 bec ause the y lack the enzyme ne eded for its synthe sis, it is the main lipox ygenase p roduct in ne utrophils, and acts as a pote nt stimulus for le ukocyte infiltration and subse que nt degranulation. 77

Prostaglandins and thromboxanes
The prostaglandins PGD 2 and PGE 2 are de te c te d at inc re ased leve ls in nasal lavage fluid follow ing allergen challe nge in subjects w ith SAR 7 8,79 and PAR, 80 but only in the e arly resp onse and not the late phase of inflammation. 6 6 Inhibitors of cycloox yge nase, the enzyme re quired for the synthe sis of p rostaglandins, do not affe ct the re sponse to antigen in the human nasal airw ay. 6 6 PGE 2 is synthe sise d by the airw ay epithe lium and has bee n proposed as a possible EpDRF. In the low e r airw ays, AHR may re sult from epithe lial damage, reduc ing PGE 2 gene ration by epithe lial c ells and, therefore, de creasing its re lax ant effect on airw ay smooth muscle . 81 How eve r, in the human nose , any action of PGE 2 w ould presumably have to be on blood ve sse ls (since there is little airw ay smooth muscle in the human nose ), and one w ould also ex pe ct a dec rease in PGE 2 release if it w as involved in AHR, yet the opposite is true. Thromboxane A 2 (Tx A 2 ) doe s appear to me diate AHR in animal mode ls 82 ,8 3 and may do likew is e in man. 84 Inhibition of thrombox ane synthesis reduce s AHR and also inhibits airw ay e osinophilia afte r alle rgen challe nge. 8 5 The thrombox anes may, the refore , have an important role in upre gulating the eosinophil-associate d response. 7 7 The contribution of thrombox ane s to nasal allergy re mains undefined, and more research is ne eded to investigate their pote ntial role in the development of AHR.

Neuropeptides
Sensory ne rve fibres c ontain a numbe r of differe nt pep tide s, including calcitonin ge ne-re late d p eptide (CGRP) and the tachykinins substance P (sub P) and ne urokinin A (NK-A). The se ne urope ptides, metabolised inte r a lia by the e nzyme ne utral endopeptidase (NEP), are re leased from se nsory nerves w hich form part of the non-adrene rgic non-cholinergic (NANC) ne rvous syste m, and are capable of gene rating a loc al ax on reflex w hich c auses an inc rease in vasc ular permeability, plasma leakage and subsequent tissue oedema. This response is know n as neuroge nic inflammation, and is me diate d by the tachykinin NK-1 and NK-2 re ce ptors. In addition, e osinop hils are capable of producing vasoac tive inte stinal p eptide (VIP) and sub P. 8 6 All these ne urope ptides are found in nasal sec re tions follow ing nasal challenge w ith grass polle n, 87 but the ir role in AHR remains uncle ar. How e ver, the re is a correlation be tw een the presenc e of AHR and the ac tivity of NEP in the human nose. 2 7 Furthe rmore , in both PAR and SAR, AHR to bradykinin is me diated by ne ural re flex es, 1 5 w hich could conce ivably include a role for neuropeptide s.
Ne uropep tide s appear to be imp ortant in AHR in a variety of animal mode ls. In the low er airw ay of the rat, AHR induced by eosinophil-derived cationic prote ins is inhibited by neurokinin NK-1 rece ptor antagonists and c apsaic in. 45 In the guinea pig, application of capsaicin to the low er airw ay (an action w hich de ple te s sensory nerve s of neurope ptides) also abolishes antigen-induc ed AHR. 88 ,8 9 The tachykinins, particularly sub P, may enhance eosinophil rec ruitment 90 ,91 and, there fore, cause AHR via an eosinophil-depe ndent mechanism. How e ver, the inhibition of AHR by c apsaic in does not affec t lipox ygenase activity or e osinop hil infiltration, suggesting that neurop eptides cause AHR indepe nde ntly of, or afte r, eosinophil activation and le ukotrie ne synthesis. 89 Ne urogenic inflammation may be a phenomenon only found in animals, since one study found no e vide nce of c apsaic in-induc ed neuroge nic inflammation in the human nasal airw ay. 92 How e ver, other studies, some utilising higher doses of capsaicin, confirm that neurogenic inflammation does oc cur in alle rgic rhinitis. 93 ,9 4 Furthermore, application of capsaicin re duc es the symptoms caused by antigen challe nge in SAR 95 and PAR 96 , though ne ithe r of these studie s investigated AHR. It is , there fore, possible that AHR in the human nasal airw ay may result, at le ast in part, from an upregulation of neuroge nic inflammation, possibly due to epithelial damage inc re asing the ex p osure of se nsory nerves. Bradykinin c auses neurope ptide re le ase in vivo in the human nasal mucosa, 97 w hile histamine has a similar effect on human lung tissue e x vivo . 98 Damage to the airw ay epithelium may inc rease ex posure of se nsory ne rves to stimuli, e ffec tively causing an upregulation of ne urogenic inflammation. This w ould result in increase d neuropeptide re lease, w hich might further pote ntiate AHR, perhaps by dire ct e ffec ts on inflammatory ce lls (e.g. e osinophils) or by stimulating cytokine release from c ells in the nasal muc osa, a proce ss w hich occurs in the human nasal airw ay. 99 Alte rnative ly, neurope ptides may not just upre gulate the inflammation re sulting in AHR, but also dire ctly c ause the inc re ased re sponse observe d in AHR. A number of stimuli, including antige n, cause ne uropep tide release in the human nasal airw ay. If tachykinin ac tivity is pote ntiated, due to more neuropeptide rele ase or inhibition of me tabolism, this w ould cause an inc rease in neurop eptide-me diate d symptoms such as nasal obstruction and rhinorrhoe a: in effect the 'hype r'-response observe d in AHR. Phosphoramidon, a NEP inhibitor, c an cause AHR in the low er airw ays of guinea pigs 1 00 and man, 10 1 and p otentiate s the ac tivity of neuropeptides in the human nasal airw ay. 10 2 NEP is pre sent in the human nasal muc osa, 1 03 and correlates w ith the presenc e of nasal AHR. 2 7 Ep ithe lial damage may cause AHR by re ducing the activity of NEP associate d w ith airw ay e pithelial cells. How e ve r, in the human nasal airw ay, PAF-induce d AHR is not associate d w ith the release of sub P (Fig. 2). PAF may cause AHR by a different me chanism to antigen, the forme r being inde pendent of neuropeptide re lease. Alternatively, PAF may pote ntiate subsequent neuropeptide re le ase from sensory nerve s, and not cause neuropeptide re le ase p e r s e , so an inc re ase in sub P w ould only occ ur in the prese nce of a further stimulus, such as histamine .
It is unknow n w hich particular neurope ptides may be involved in the de ve lopme nt of AHR. Sub P is the most pote nt e ndogenous me diator at the NK-1 rec eptor, but it doe s not induc e AHR in the low er airw ays of shee p, w hile NK-A doe s. 10 4 In the human nasal airw ay, NK-1 rec eptors are localised to the epithe lium, glands and blood ve ssels, w hile NK-2 rec eptors are limite d to arterial vessels. 105 This is in agre eme nt w ith the findings that nasal obstruc tion is me diated through the NK-1 rece ptor, w hile plasma ex travasation probably involve s both NK-1 and NK-2 rec eptor activation, 1 06 and w ould ex plain the inconsis te nc y that NK-1 or NK-2 re ce ptor antagonists preve nt AHR in some studie s but not in others. In chronic asthma, both rec eptor subtypes are upre gulated, 10 7 and this may happen in chronic allergic rhinitis as w ell. The re are no re ports investigating the role of CGRP in AHR.

Cytokines
Cytokine s are inte rc ellular me sse nger pe ptides w hich are rele ase d by a variety of c ells to influe nce the activity of othe r ce lls. Three cytokines are of vital importance in the development and regulation of eosinophil func tion: the interle ukins IL-3 and IL-5, and granulocyte-macrophage colony-stimulating factor (GM-CSF). All thre e p re vent ap optosis and prolong the survival of eosinophils in vitro , and in partic ular, IL-5 is essential for the diffe rentiation of proge nitor cells into eosinophils. 1 08 Both IL-4 and IL-5 have bee n implic ated in the development of AHR. 10 9 In animals, IL-5 c ause s a marked e osinophilia, e osinophil ac tivation and AHR. 9 1 Monoclonal antibodies to IL-5 abolish antigen-induce d eosinophilia and AHR in the low er airw ay of the guinea p ig. 1 10 IL-4 regulate s the ac tivity of CD4+ T-lymphocyte s, w hich re lease a range of cytokine s capable of priming and activating e osinop hils, 71 and can also ac tivate neutrophils. 1 11 Furthe rmore , memory T-ce lls in the nasal mucosa of patients w ith nasal alle rgy can p roduce IL-4 during allergen ex posure, w hich may upre gulate the inflammatory response . 11 2 Patients w ith SAR or PAR have a raise d numbe r of CD4+ T-c ells. 11 3 Follow ing nasal allerge n challe nge, the levels of IL-1a , IL-1b , IL-5, IL-6, IL-8 and GM-CSF are raise d in nasal sec re tions, 1 14 -11 6 and human eosinophils are pote ntial source s of these cytokines. 1 17 Epithelial ce lls, is olated from allergic rhinitics, show e d inc re ased immunostaining for GM-CSF, IL-8, the rec eptors for IL-1 and TNF-a , 11 8 and also the y rele ase more IL-1b , IL-8, GM-CSF and TNF-a compared to epithelial ce lls from non-alle rgic subjects. 1 19 Similar inc reases in IL-4-, IL-5-and GM-CSF-positive cells are observe d in biopsie s from the nasal muc osa of atopics. 12 0 Both inte rferon-g and TNF-a (and possibly other cytokines) cause an upre gulation of ICAM-1 on human nasal epithe lial cells, 1 21 w hile IL-4 upregulates the ex pression of VCAM-1. 11 3 Both these adhesion mole cule are upregulate d in alle rgic rhinitis. 113 Cytokines may promote AHR by upregulating the rec ruitment and activation of eosinophils (Fig. 3) and neutrophils. For ex amp le , there are close correlations betw e en the numbe r of eosinophils and GM-CSF le ve ls in bronchoalve olar fluid (BALF) from subjects w ith asthma, 1 22 w hile the survival of eosinophils in BALF from subje cts w ith alle rgic rhinitis correlates the c onc entrations of IL-5 and GM-CSF present. 1 23 Ne urokinin NK-2 re cep tor antagonists inhibit AHR induced by IL-5 in the guinea pig, but not the associate d e osinophilia, 1 24 indic ating that cytokines are involved in ce ll rec ruitment, follow ing w hich various me diators are re le ased, such as neuropeptides w hich may c ause the development of AHR. IL-1 may also be involved in the deve lopme nt of AHR, 12 5 but it has a w ide range of c ellular ac tions, so it is difficult to sugge st a p re cise role for it in the pathogenesis of AHR. Intere stingly, it may inhibit the activity of NEP, 12 6 or induce the gene ration of ox ygen-base d fre e radic als from mac rophages and ne utrophils. 12 7 Finally, glucocorticoids dow nre gulate the production of IL-3 and IL-5 follow ing alle rgen challenge; this effect may c ontribute to the action of ste roids in abolishing AHR, 11 6 although glucoc ortic oids have othe r ac tions too.

Chemokines
Che mokines are c ytokine s w hich posse ss chemotac tic activity, and are divided into groups de pending on their structure. The tw o main groups are CC chemokine s, w here tw o cysteine re sidue s are adjac ent to each other (e.g. RANTES, MIP-1a , eotax in) and CXC che mokines, in w hich the tw o c yste ine re sidue s are se parated by a third amino acid (e.g. IL-8). Some che mokines (RANTES, eotax in) appe ar to be selec tive for eosinophils, w hile IL-8 has a che motactic ac tivity only for neutrophils (though there are re ports that it also has an action on prime d human eosinophils). 1 28 The levels of RANTES, 1 15 ,1 29 MIP-1a , 1 15 eotax in 13 0 and IL-8 11 5,13 1 dete cted in nasal lavage are raise d follow ing nasal allerge n challenge in man. Treatme nt w ith gluc oc ortic oids, w hich inhibits inflammatory ce ll re cruitme nt, abrogate s these incre ase s. 1 15 ,13 2 Interestingly, mucosal c ells obtaine d from the nose s of subje cts w ith allergic rhinitis show increase d ex pre ssion of mRNA for RANTES 1 29 and e otax in. 1 30 It is now ge nerally acc epted that RANTES and eotax in are important in IL-5-mediate d eosinophilia, w here the latte r c ause s the mobilisation of e osinophils into the circ ulation w hile the local re lease of chemokines provide s a 'homing' mechansim for the migration of eosinophils into the tissues. 12 8 How ever, no study has ye t investigated the spec ific role s of che mokine s in nasal hyperre sponsivene ss. Notw ithstanding, administration of RANTES into the nasal airw ay of subjec ts w ith alle rgic rhinitis causes an e osinophilia but not an influx of other inflammatory c ells. 1 33 How ever, the same study also found that, after allerge n challe nge, administration of RANTES also caused an influx of basophils, neutrophils, lymphocytes and monocytes, as w ell as c ausing epithelial she dding, a response similar to that obse rved in nasal hyperre sponsivene ss. It is there fore likely that chemokines have an important role in the recruitme nt of inflammatory c ells that is obse rved during the de ve lopme nt of nasal hyperresp onsive ness.

Nitric oxide
Rec ently, a large amount of rese arch has bee n carrie d out on the role of nitric ox ide in airw ay disease . There is e videnc e that nitric ox ide synthase (NOS) activity is inc re ased in PAR 134 and in SAR. 13 5,13 6 This w ould inc re ase the amount of nitric ox ide available to react w ith supe rox ide , gene rating free radicals w hich may contribute to the de ve lopme nt of AHR. 33 Nitric ox ide may also have a role in the production of c ytokine s ne cessary for e osinophil survival, such as IL-4 and IL-5. 13 7 Therefore, one might ex p ect inhibitors of NOS to pre vent AHR, either by re duc ing free radical production or via the induc tion of e osinop hil ap optosis . Paradox ically, NOS inhibitors actually induc e AHR to histamine and bradykinin in both the human nasal airw ay 1 38 (Fig. 4) and low er airw ays of asthmatic s, 13 9 and also in animal mode ls. 14 0 It is possible that in the normal, non-inflamed airw ay, nitric ox ide is prote ctive, so inhibitors of NOS w ould c ause AHR, but in chronic airw ay inflammation, an inapp ropriate degre e of nitric ox ide produc tion may be harmful and pote ntiate the allergic response.
Intere stingly, NO is thought to be the ce ntral me diator of inhibitory NANC transmission. There fore, inhibition of NOS may cause a reduction in the activity of inhibitory NANC ne rves, w hich could cause AHR by potentiating ne urogenic inflammation me diated by ex citatory NANC ne rves. In chronic alle rgy, ex cessive NO production could c ause AHR by the formation of the perox ynitrite fre e radical, w hich causes AHR in the guinea p ig airw ay, possibly by inhibiting cGMP production. 33 Furthermore, other NO metabolite s, such as nitryl chloride, can be synthesised by ne utrophils, inactivating e ndothelial ce ll angiote nsin-converting e nzyme . 141 This enzyme is involved in the degradation of kinins and p ossibly tachykinins in allergic rhinitis, 10 2,1 42 so inhibition of this e nzyme may influe nce AHR by pote ntiating the action of these mediators.

Kinins
A numbe r of studie s have identified a role for kinins in the de ve lopme nt of AHR in the low er airw ays of both the guinea pig 14 3 and she ep. 1 44 Furthermore, the ability of MBP and synthe tic c ationic p rote ins to induc e AHR in the low er airw ays of rats is depe nde nt on the gene ration of kinins. 4 6 In all three studies, administration of a bradykinin B2 re cep tor antagonist inhibited the de ve lopme nt of AHR, and, w he re investigate d, also appeared to have an effect on the rec ruitment of eosinophils into the airw ay. ECP c an stimulate kallikre in activity in v itro , 14 5 re sulting in kinin production, so e osinophil-de rived cationic p rote ins may generate kinins w hich lead to AHR.
Re ce nt investigations have found that AHR in the human nasal airw ay may also be kinin depe nde nt. Ic atibant, a highly pote nt antagonist at the bradykinin B2 rec eptor, prevents PAF-induce d AHR (Fig. 5), w hile PAF c ause s an incre ase in the le vel of kinins in nasal lavage fluid (Fig. 6). Kinins are produc ed in both PAR and SAR, and could therefore contribute to AHR in alle rgic rhinitis.
One important c onside ration is the source of the kinins. In SAR, antige n challe nge causes an inc re ase in plasma ex travasation, so the kinins dete cte d may be a product of plasma kallikre in activity. How ever, antigen doe s not induc e plasma ex travasec tion in PAR, 5 and the role of plasma kallikrein has not bee n assessed in either alle rgic rhinitis or asthma. 1 46 The le ve ls of tissue kallikrein in nasal sec retions are inc re ased in allergic rhinitis , 1 47 but it is not thought that tissue kallikrein can be activated by cationic prote ins. 4 6 A novel solution to this anomaly is as follow s. Neutrophils contain tissue kallikre in and bind plasma kallikrein, together w ith high-and low -molec ular w e ight kininogen, on the ce ll surfac e. 1 48 ,14 9 Antigen and PAF challe nge c ause a ne utrophilia in addition to e osinophil re cruitment, 54 therefore neutrophils may provide the c ompone nts nee de d for the gene ration of kinins.
The mechanism by w hich kinins induce AHR remains unknow n. Ex ogenous kinins applied to the airw ays do not c ause AHR in animal models 4 6 nor in the human nasal airw ays (Turner & Foreman, unpublishe d data), though one study found that bradykinin can produce AHR to ace tylcholine in the guinea pig airw ay. 1 50 Bradykinin c auses se nsitisation of C-fibres in the guine a pig trachea, 1 51 and the re is e videnc e that in the human nose e nhanc ed resp onsive ness to bradykinin is mediated by neural reflex e s. 15 Bradykinin c an also re lease sub P and other ne urope ptides from se nsory nerve endings, 98,15 2 so it may induce AHR by a ne uropeptide -de pendent mechanism. Howe ve r, although administration of PAF into the human nose causes kinin ge neration, there is no increase in sub P release . Alternatively, bradykinin c an initiate the production of the c ytokines IL-1, IL-6 and IL-8 in v ivo 1 53 and stimulate the re lease of TNFa /b and IL-1 from macrophage s. 15 4 These c ytokine s may contribute to AHR as pre viously mentione d. Inte restingly, the eosinophilia induc ed by antigen challenge in the low er airw ay of the guine a p ig is reduc ed by bradykinin rec eptor antagonists 14 3 or an inhibitor of tissue kallikre in, 15 5 thus providing e videnc e for the involveme nt of kinins in the recruitme nt of e osinophils follow ing ex posure to allerge n.

The role of inflammatory cells in hyperresponsiveness
From the evidence pre se nted, it might appe ar that eosinophils have a vital role in the deve lopme nt of AHR. How e ver, the re lationship be tw een eosinophil activation and AHR remains c ontroversial. Ce rtainly, eosinophils are involve d in the late response , but the development of AHR may not be dep endent on the presenc e of eosinophils. Studie s c onducted in animal mode ls indicate that AHR can occ ur w ithout a dete ctable eosinophilia, and vice versa. 1 56 Data obtained from studie s in the human nose similarly imply that AHR doe s not ne cessarily occ ur toge ther w ith e osinophil ac tivation, 48 ,49 w hile antige n can induc e an e osinop hilia w ithout c ausing AHR. 1 57 There are a range of airw ay inf lammatory c onditions w hich feature an eosinophilic infiltration, but no assoc iate d AHR. 22 None the less, the close association betw e en eosinophil recruitme nt, ac tivation and AHR in many studie s 35 ,52 ,67 implies, at the very least, that eosinophils may contribute to the deve lopme nt of AHR.
Although e osinophils are the main ce lls w hich have bee n imp lic ated in the me chanism underlying AHR, it is like ly that other c ells are involve d, including ne utrophils, T-lymphoc ytes, macrophage s, B-cells and basophils. 7 1 Basophils, w hich may be involve d in the late alle rgic re sponse , are capable of gene rating histamine, bradykinin, MBP, LTC 4 , IL-4 and IL-8, 71 ,15 8 w hich may contribute to AHR. Macrophage s may also be involved in the de velopment of nasal AHR, possibly by generating c ytokine s and free radicals. 1 59 The involvement of the ne utrophil in AHR is often overlooke d, yet there is a w ealth of evide nce for an active role. Not only may the y c ontribute to the generation of kinins and cause tissue damage via the supe rox ide burst, but the y also ge nerate PAF, LTB 4 and a variety of c ytokine s. It is interesting to note that ECP has now been detected in ne utrop hils. 16 0 Neutrophils isolate d from atopic subje cts have an inc re ased c ap ac ity for myelope rox idase re le ase, and this enzyme c an ge nerate nitric ox ide-derived ox idants, w hich c an induc e AHR and inhibit kinina se II, an enzyme responsible for the metabolism of kinins . 1 41 The hypothesis that both ne utrophils and eosinophils ac t together in c ausing AHR is one w hich w arrants furthe r inve stigation.
The upregulation of ICAM-1 and VCAM-1 by cytokines in allergic rhinitis has alre ady bee n disc usse d. ICAM-1 and VCAM-1, present on endothe lial cells, bind respe ctive ly to integrins LFA-1 (a Lb 2 integrin) and VLA-4 (a 4b 1 inte grin) on leukoc yte s, allow ing leukoc yte s to adhere to the endothe lium prior to migration. The inte raction be tw een VCAM-1 and VLA-4 appe ars to be particularly important in eosinophil migration, as neutrophils do not have the VLA-4 rec eptor on their surfac e. 1 61 Interfe ring w ith the VCAM-1/VLA-4 pathw ay may, there fore, prevent eosinophil-depe nde nt hyperre sponsivene ss. Monoclonal antibodie s (mAbs) to the integrin sub-unit a 4 have bee n developed w hic h abolish antigen-induce d hyperresp onsive ness in the low er airw ays of a number of animal mode ls of airw ay alle rgy. 1 62 -1 64 Inte restingly, this e ffec t w as not alw ays associated w ith an decre ase in the airw ay eosinophilia, 1 63 ,1 65 possibly bec ause other me chanisms ex ist (e .g. via ICAM-1) w hereby e osinophil migration may oc cur. Further development of these mAbs may provide a useful therapeutic intervention for airw ay alle rgy in man.

Summary
It w ould appear that a number of me diators are capable of inducing AHR in the human nasal airw ay, perhaps acting via different mechanisms, w hich w ould help ex p lain the c onflic ting e videnc e regarding a role for eosinophils or airw ay e pithe lial damage in this proce ss. How eve r, the fac t that inhibitors of leukotriene s, PAF, kinins and tachykinins can all inhibit antige n-induced AHR implie s that there may be a ce ntral common pathw ay in the pathogenesis of AHR. Sinc e neuroge nic inflammation c an pote ntially be modulate d by all of the se mediators, neuropeptides may be involved in the final stage of AHR induc tion, perhaps as show n in Fig. 7. Antagonists of these me diators appe ar to inhibit the de velop ment of hyperresp onsive ness in animals, and the de ve lopme nt of similar antagonists for use in man may provide new pharmacologic al treatments for allergic rhinitis, and perhaps for other alle rgic airw ay dise ases as w e ll.