Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The structurally related tachykinins (or neurokinins), substance P (SP), neurokinin A (NKA), and neurokinin B (NKB), are widely distributed in the central and peripheral nervous systems. Biological effects of these neuropeptides are carried out via binding to their preferred receptors, NK₁, NK₂, and NK₃, respectively, which are members of the G protein-coupled receptor superfamily. Activation of the tachykinin receptors influences a broad array of biological actions, including contraction, secretion, immune responses, and neurotransmission. The functions of a number of tissues (cardiovascular, respiratory, digestive, reproductive, excretory, and musculoskeletal) are influenced by these neuropeptides (Otsuka and Yoshioka, 1993).

SP and NKA are encoded from a single gene (preprotachykinin A or PPT-A), which gives rise to multiple mRNAs, (, , and ) through RNA splicing (Krause et al., 1987). The amounts of these precursors are regulated in a tissue-specific manner; this determines, in part, the local levels of the neuropeptides (Otsuka and Yoshioka, 1993). Coexpression of SP and NKA exists in many tissues: human skin (Schulze et al., 1997), human bladder (Smet et al., 1997), rodent tendons, and joint capsule (Ackermann et al., 1999) and guinea pig airways (Kummer et al., 1992). The NKB gene (PPT-B) shows structural similarity to PPT-A; however, the preprotachykinin A and B mRNAs differ in the major sites of their expression (Kotani et al., 1986). Recently, SP and NKB were both detected in rodent ileum (Yunker et al., 1999). Up-regulation of the PPT genes and mRNAs for the neurokinin receptors occurs both in animal models of disease, such as allergic inflammation of the lungs (Fischer et al., 1996) and in human diseases, such as asthma (Adcock et al., 1993; Bai et al., 1995).

Development of structurally diverse tachykinin receptor-selective antagonists has advanced our understanding of the biological actions of these neuropeptides. Clinical studies have shown that these antagonists (particularly NK₁ receptor antagonists) offer potential management of emesis (Navari et al., 1999) and other illnesses such as major depression (Kramer et al., 1998). However, NKA and NKB also bind with moderate affinity to the NK₁ receptor (Maggi and Schwartz, 1997), possibly at a site distinct from the SP binding domain (Wijkhuisen et al., 1999). This nonselective nature of tachykinin binding, as well as the colocalization of these neuropeptides within tissue, suggests that some diseases, in particular asthma, might benefit from blockade of more than one tachykinin receptor. For example, Turner et al. (1996) have reported that airway hyper-reactivity and inflammation in nonhuman primates were synergistically improved by treatment with both NK₁ and NK₂ receptor antagonists compared with separate treatment with either compound alone. Thus, we have developed a series of compounds with affinity for all of the three tachykinin receptors. This report provides the pharmacological characterization of a series example, 3-cyano-N-((2S)-2-(3,4-dichlorophenyl)-4-[4-[2-(methyl-(S)-sulfinyl)phenyl]piperidino]butyl)-N-methyl-]-napthamide (ZD6021). The data indicate that ZD6021 is orally available and is a potent antagonist of all three tachykinin receptors. Future studies of tachykinin-related pathology such as asthma may benefit from this pharmacological tool.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Binding Studies. The cloning, heterologous expression and scale-up growth of MEL cells transfected with either the NK₁, NK₂, or NK₃ receptor were conducted as previously published for the human NK₂ receptor (Graham et al., 1991; Takeda et al., 1991; Huang et al., 1992; Aharony et al., 1994). The human NK₁ receptor was identical to that reported previously (Gerard et al., 1991; Fong et al., 1992), whereas the human NK₃ receptor differed from the genomic sequence at AA439 (Cys versus Phe; Buell et al., 1992; Takahashi et al., 1992).

Mucus Secretion Assay. A Chinese hamster ovary-K1 cell line transfected with the human NK₁ receptor was used to determine the ability of ZD6021 to block mucus secretion (Caccese et al., 1999). These cells were used prior to reaching confluency and labeled with [³H]glucosamine. The cells were then stimulated with the NK₁ receptor agonist Ac-[Arg⁶, Sar⁹, Met(O₂)¹¹] substance P_6-11 (ASM-SP) in the presence or absence of ZD6021. The medium was collected, refrigerated for mucin determination, and subsequently precipitated with 10% trichloroacetic acid:1% phosphotungstic acid. The precipitate was centrifuged, washed twice in 10% trichloroacetic acid:1% phosphotungstic acid, and redissolved in 0.1 N NaOH. The labeled glycoconjugates were counted (Packard Tri-Carb 2200CA, Downers Grove, IL). A secretory index was calculated by dividing the counts from the stimulation period by the counts from the previous equilibration period. Triplicate measurements were obtained for at least three experiments to derive an average secretory index. The IC₅₀ was determined by curve-fitting the mean data in a nonlinear regression analysis using a sigmoidal dose-response with a variable slope and expressed as the log IC₅₀ (pIC₅₀) value.

Isolated Tissue Responses. Male New Zealand White rabbits (2-3 kg) were administered heparin (2.5 U/kg) and sodium pentobarbital (60 mg/kg i.v.). A bilateral thoracotomy was performed and the left and right branches of the pulmonary artery were excised, trimmed free of connective tissue, and cut into 5-mm rings. In some cases, the endothelium was removed by gentle abrasion of the intimal surface. The segments were suspended in water-jacketed tissue baths, thermostatted at 37°C and containing physiological salt solution comprised of 119 mM NaCl, 4.6 mM KCl, 1.8 mM CaCl₂, 0.5 mM MgCl₂, 1 mM NaH₂PO₄, 25 mM NaHCO₃, and 11 mM glucose. The medium was gassed continuously with O₂:CO₂ (95:5). Initial tension was set at 2 g and stabilized for 30 min. Changes in tension were monitored by force transducers (Grass FT-03, Quincy, MA) coupled to a data acquisition system (Grass polygraph, model 7, interfaced with MI²; Modular Instruments, Malvern, PA).

Tracheal Extravasation. A modified method of Saria et al. (1983) was used with Evans blue dye as an indicator of protein extravasation. Male guinea pigs (250-400 g) were anesthetized (1.5 g/kg i.p.) and surgically implanted with a jugular catheter for administration of compounds. All animals were pretreated (i.v.) with indomethacin (10 mg/kg) and propranolol (0.5 mg/kg) to reduce the indirect influence on pulmonary responses of cyclooxygenase products or -adrenergic receptor activation (Buckner et al., 1991). Thiorphan (10 mg/kg i.v.) was added to prevent possible degradation of the tachykinin receptor agonist and to potentiate its responses (Kusner et al., 1992). The animals were also pretreated with the bradykinin₂ receptor antagonist HOE-140 (0.1 µmol/kg i.v.). In preliminary experiments, the additional administration of HOE-140 reduced the basal level of extruded dye from 48 ± 12 (n = 5) to 8.0 ± 1 (n = 6) of Evans blue dye per milligram of trachea. Evans blue dye (30 mg/kg) was intravenously administered 10 min following the final pharmacophore addition and 1 min prior to administration of ASM-SP (0.1 nmol/kg i.v. or ED₉₀) or its saline vehicle. To remove intravascular dye, perfusion of the animal via the aorta was performed with saline at a rate of 60 ml/min × 2 min. A 1-cm length of the trachea, rostral to the bifurcation, was excised, weighed, and placed in 2 ml of 100% formamide at 60°C for 18 h. Triplicate values of the absorbance at 620 nm (Molecular Devices Spectramax 250, Sunnyvale, CA) of the formamide/dye solution were used to calculate the level of extravasation, expressed as nanograms of Evans blue dye per milligram of trachea. Protection was expressed as a percentage of the difference in extravasation produced in the absence or presence of agonist (ED₉₀ dose of ASM-SP, 0.1 nmol/kg i.v.).

Airway Mechanics. Measurement of airway function in male albino guinea pigs (300-600 g) was made in the absence or presence of ZD6021 according to previously described methods (Buckner et al., 1993). Anesthetized (urethane, 1.5 g/kg i.p.) guinea pigs were surgically instrumented for administration of compounds (jugular vein) and recording of blood pressure (carotid artery). The trachea was cannulated and connected to a heated pneumotachograph (Fleisch 0000; Switzerland), which was also attached to a Validyne differential pressure transducer (model MP45-14). Transpulmonary pressure was obtained by measuring the pressure difference between an intrapleural cannula placed in the 5th intercostal space and a sidearm adapter of the tracheal cannula with a second Validyne differential pressure transducer (model 45-24). Pulmonary resistance (R_P) and dynamic lung compliance were calculated (Modular Instruments). The animals were pretreated intravenously with ancillary pharmacophores (indomethacin, propranolol, and thiophan) as described above.

Statistical Analyses. Data were expressed as mean ± S.E.M. Statistical differences were determined using ANOVA/Tukey-Kramer or Student's t test, with a minimum significance of p < 0.05.

Chemicals. The tachykinin antagonist ZD6021 (Fig. 1) was prepared as the crystalline citrate salt. This compound and other tachykinin antagonists (SR 140333, ZD7944, ZM274773) were synthesized in AstraZeneca chemistry research laboratories. For in vitro experiments, ZD6021 was dissolved in dimethyl sulfoxide. For in vivo work, it was dissolved in 30% polyethylene glycol 400 balanced in saline. Thiorphan, indomethacin, and propranolol were purchased from Sigma (St. Louis, MO). Senktide was obtained from Penninsula Laboratories (San Carlos, CA) and ASM-SP and BANK from Cambridge Research Laboratories (Cheshire, UK).

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Structure of ZD6021. ZD6021 was formulated as the crystalline citrate salt.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

In Vitro Pharmacology. Figure 2 shows that affinity of ZD6021 for the human tachykinin receptors was high. The binding of either [³H]SP or [³H]NKA to its putative receptor was potently inhibited by ZD6021: K_i = 0.12 ± 0.1 nM (n = 10, pK_i = 9.95 ± 0.04) and 0.61 ± 0.7 nM (n = 6, pK_i = 9.23 ± 0.05), respectively. Comparatively, ZD6021 inhibition of ligand (¹²⁵I-MePhe7NKB) binding at the NK₃ receptor was markedly lower: K_i = 74.0 ± 0.7 nM (n = 3, pK_i = 7.13 ± 0.04). Binding affinity of ZD6021 for nontachykinin receptors (or sites) with K_i values less than 1 µM was limited to the dopamine D_2L receptor (90 nM), the muscarinic M₂ (509 nM) receptor, the dihydropyridine L-type calcium channel (606 nM), and the sigma nonselective site (777 nM). Although binding affinity at the muscarinic M₃ receptor was low (K_i = 1.2 µM), it was the only nontachykinin site that showed marked antagonism (>50%) of receptor activation by ZD6021 (10 µM) in an isolated tissue assay (guinea pig ileum).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Inhibition of ligand binding to cloned human tachykinin receptors by ZD6021. MEL cell membranes with human recombinant NK1 (squares), NK2 (circles), and NK3 (triangles) receptors were incubated with ~1 nM of the indicated ligand in the presence of increasing [ZD6021]. Nonspecific binding was determined with 1 µM of the homolog ligand. Values, expressed as percentage of control in duplicate, represent means ± S.E.M. (n = 3-10).

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Dose-dependent antagonism of NK1- and NK2-mediated changes in contractile activity of rabbit pulmonary artery. Effect of ZD6021 on concentration-response effects to ASM-SP (panel A) or BANK (panel B) in isolated rabbit pulmonary artery. Contraction of the pulmonary artery was stimulated by BANK in tissues denuded of endothelium. Values are means ± S.E.M. and are expressed as a percentage of the maximum response to either 100 µM papaverine for ASM-SP-induced relaxation responses or 30 mM BaCl2 for BANK-stimulated contraction. The slopes from Schild regression (insets) were not significantly different from unity.

8

5

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Effect of ZD6021 on senktide-induced contraction of guinea pig ileum. Longitudinal segments of the ileum were prepared as described under Experimental Procedures. Responses to senktide in the absence and presence of ZD6021 (0.01-10 µM) were referenced to a maximal contraction elicited by 1 mM BaCl2. Values represent means ± S.E.M. (n = 6-9).

In Vivo Pharmacology. The blockade of NK₁ or NK₂ receptor activation by ZD6021 was evaluated by determining its ability to prevent tracheal extravasation of Evans blue dye (NK₁ only) and changes in airway mechanics (NK₁ and NK₂). For the former assay, a single agonist dose (ASM-SP = 0.1 nmol/kg i.v.) was given to each animal that had been treated either with vehicle or ZD6021. In animals given the vehicle, this dose resulted in near maximal levels of extravasation, i.e., ED₉₀. The agonist-mediated response was dose dependently inhibited by ZD6021 (Fig. 5A). Half-maximal protection (ED₅₀) was determined to be at doses of 0.005 µmol/kg i.v. and 0.83 µmol/kg p.o. The feeding state of the animals did not markedly influence the level of protection offered by ZD6021. After an 18-h fast, administration of ZD6021 (1 µmol/kg p.o.) resulted in a level of protection similar to that found in animals permitted food ad libitum; i.e., 62.8 ± 10.1% in fasted animals versus 74.9 ± 5.6% in fed ones (n = 7 for both groups).

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Dose-dependent antagonism of ZD6021 in NK1- and NK2 receptor-mediated airway events. Animals were fasted overnight and treated (2-h pretreatment time) as described under Experimental Procedures. For both ASM-SP-induced plasma extravasation (panel A) and BANK-mediated bronchoconstriction (panel B), the dose of agonist was 0.1 nmol/kg i.v. (equal to ED90 in control animals). Protection in panel A was expressed as a percentage of the difference in extravasation produced in the absence or presence of agonist. Significant differences (*p < 0.05) were obtained in A between the low dose (0.3 µmol/kg p.o.) and each of the higher doses (0.6, 1, and 3 µmol/kg p.o.). In B, the abbreviation GL (y-axis) refers to airway conductance (or 1/Rp). Each animal received cumulative incremental doses of agonist, producing progressive declines in GL compared with the initial value. The dose of ZD6021 that resulted in only a 50% decrease in conductance after receiving 0.1 nmol/kg i.v. BANK was defined as the ED50. In animals that did not receive BANK, there was a 10% loss in conductance of over the time course of the experiment. Thus, in animals treated with ZD6021, conductance was not maintained at 100% of the initial baseline value. Significant differences (*p < 0.05) within treatment groups were obtained between either 10 or 20 µmol/kg p.o. and 30 and 100 µmol/kg p.o. Values represent means ± S. E. M (n = 3-10).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Time-dependent antagonism by ZD6021. Animals received a single oral dose of ZD6021; in A, 3 µmol/kg versus ASM-SP-induced (0.1 nmol/kg i.v.) tracheal plasma extravasation, and in B, 30 µmol/kg versus BANK-mediated (0.1 nmol/kg i.v.) bronchoconstriction. The pharmacodynamic half-life of ZD6021, i.e., the time point when protection had decreased by 50%, was calculated by linear regression analysis of the decay portion of the curve, beginning with the value coinciding with the maximum level of protection and immediately prior to a fall in protection. The time to reach maximum protection was then subtracted. Refer to Fig. 5 for definition of the term "protection" in panel A. In panel B, protection by ZD6021 was expressed as a percentage of the difference between the conductance obtained from the ZD6021-treated group and that found for the control animals divided by the difference between the maximal conductance value and the control group conductance value. All values represent means ± S.E.M. (n = 3-6/time point).

View larger version (26K): [in this window] [in a new window]   Fig. 7.   Effects of NK1 and NK2 receptor blockade on capsaicin-mediated bronchoconstriction. Anesthetized guinea pigs were treated intravenously with cumulative doses of capsaicin in the absence or presence of antagonist. Animals were first pretreated with ancillary pharmacophores as described under Experimental Procedures. They were then given either the selective NK1 receptor antagonist SR 140333 (1 µmol/kg i.v.), or the selective NK2 receptor antagonist ZD7944 (0.3 µmol/kg i.v., A. Shenvi, D. Aharony, and D. Lengel, unpublished data), or in combination at these same doses, or ZD6021 (10 µmol/kg i.v.). Capsaicin was given in 10-min intervals, allowing full recovery of the respiratory response. Values represent means ± S.E.M. (n = 5-6/group).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The purpose of the present study was to characterize the pharmacological activity of a novel antagonist of all three tachykinin receptors. The major findings indicate that ZD6021 is a potent, orally active antagonist capable of attenuating tachykinin-mediated responses. The in vitro activity of ZD6021 was evaluated in three separate assay systems: ligand binding to the human cloned receptors, mucous secretion by cultured cells, and smooth muscle contractile function using isolated tissues. In all cases, the compound potently blocked agonist-mediated events in a dose-dependent manner. Potency values were highest from binding studies of the human cloned receptors. The K_i values for the human NK₁ and NK₂ receptors were both subnanomolar with about 5-fold greater affinity for the NK₁ receptor. The affinity constant for the human NK₃ receptor was about 2 orders of magnitude lower than that of the NK₂ receptor. Nonetheless, at appropriate concentrations of ZD6021, inhibition of all three receptors can be gained, thereby providing a useful pharmacological tool for examination of disease processes such as asthma.

Although the strong antagonist-receptor interaction was reflected in favorable response inhibition, potency values of ZD6021 were markedly reduced in some cases compared with those obtained in binding assays. For example, the pIC₅₀ obtained in the mucous secretion assay was only 7.6, whereas the pK_i was greater than 9.0 in binding studies for the NK₁ receptor. The former value was similar to that found for the selective NK₁ receptor antagonist SR 140333 in the same assay (pIC₅₀ = 8.0, Caccese et al., 1999) but this compound also showed subnanomolar affinity in binding assays (Emonds-Alt et al., 1993). Other workers have also reported discrepancies of potency values between different assays for the NK₁ receptor as well as differences in the manner of antagonism, i.e., competitive versus noncompetitive. Goldhill et al. (1999) reported competitive antagonism by SR 140333 of SP-induced secretion from colon epithelial tissue and noncompetitive blockade of SP-mediated contraction of isolated ileum. Inhibition by SR 140333 of SP binding to rat brain NK₁ receptors was also found to be competitive but inhibition of agonist-mediated relaxation of isolated rabbit pulmonary artery was noncompetitive (Emonds-Alt et al., 1993). Construction of the assay conditions, e.g., the origin of the tissue, seemed to impact the outcome of the responses and the potency values.

Since the pharmacology of the human NK₁ and NK₂ receptors measured in vitro closely resembles the biological activity found in rabbit tissues (Coge and Regoli, 1994), we chose to profile the pharmacology of ZD6021 in tissues from these species. Potency of ZD6021 for the NK₁ receptor was similar between species, consistent with earlier work. For example, the dissociation constant of the competitive NK₁ receptor antagonist CP 99994 is in the nanomolar range in both isolated human pulmonary artery (Corboz et al., 1998; Pedersen et al., 2000) and rabbit vena cava (Coge and Regoli, 1994). Potency determinations of compounds such as SR 140333, which display noncompetitive antagonism, are also comparable in tissues prepared from the two species (Emonds-Alt et al., 1993; Pedersen et al., 2000).

Unlike the potency measures of ZD6021 for the NK₁ receptor, the NK₂ receptor values obtained from human and rabbit tissues markedly differed, by an order of magnitude. These results were surprising based on previous studies. Contraction of the rabbit pulmonary artery or isolated human bronchus was competitively blocked by the selective NK₂ receptor antagonist SR 48968, resulting in pK_B values of about 9 (Advenier et al., 1992). In our hands, the potency of SR 48968 was also similar; pK_B values = 9.3 and 9.0 for rabbit pulmonary artery and human bronchus, respectively (data not shown). The isolated human pulmonary artery provides a system free of contaminating contractile effects mediated by NK₂ or NK₃ receptors (Pedersen et al., 2000), and tachykinin-induced contraction of isolated human bronchus is mediated only by NK₂ receptors (Sheldrick et al., 1995). Therefore, the reason for the disparity between the sets of results with ZD6021 for the NK₂ receptor for the two species is not readily apparent. Of note, however, are major discrepancies, i.e., 2 to 3 orders of magnitude, that have been detected between binding affinities for the human neurokinin receptors and pharmacological responses in human tissues (Pedersen et al., 2000).

The change of vascular integrity following tissue exposure to substance P has long been accepted as a valid method for evaluating potency of tachykinin antagonists. Substance P gives rise to extravasation of plasma proteins in a number of tissues (Lembeck and Holzer, 1979), and this effect has been linked to allergic airway responses (Saria et al., 1983). Changes in vascular permeability occur by the formation of gaps between endothelial cells of postcapillary venules (Baluk, 1997). Deletion of the gene encoding neutral endopeptidase results in widespread basal plasma extravasation, which can be restored upon treatment with NK₁ receptor antagonists (Lu et al., 1997). Extravasation produced by pharmacological or physiological stimuli has been blocked by pretreatment with selective NK₁ receptor antagonists. Reported values (ED₅₀) were 10 µg/kg p.o. for CP 122,721 in guinea pig lung (McLean et al., 1996), 7 µg/kg i.v. for SR 140333 in rodent skin (Emonds-Alt et al., 1993), and 150 µg/kg i.v. for RP 67580 in a similar preparation (Inoue et al., 1996). Comparatively, ZD6021 provided a similar level of blockade (ED₅₀ = 500 µg/kg p.o.) or was more active (ED₅₀ = 3 µg/kg i.v.), dependent upon the route of administration, than these selective NK₁ receptor antagonists. Moreover, the activity of ZD6021 was well sustained after oral delivery; the pharmacodynamic t_1/2 was 281 min with peak activity occurring at 150 min.

Administration of either substance P or neurokinin A by inhalation or intravenous injection produces bronchoconstriction, which can be measured as labored breathing (dyspnea) or a decrease in pulmonary function in guinea pigs (Kusner et al., 1992; Buckner et al., 1993). Pretreatment with inhibitors of neutral endopeptidase exacerbates this response in guinea pigs (Kusner et al., 1992; Savoie et al., 1995) and humans (Cheung et al., 1993). In guinea pigs, blockade of the NK₂ receptor caused the dose-response curves elicited by NK₂ receptor agonists to shift to the right. This shift was more pronounced, by as much as one log unit, in the presence of a selective NK₁ receptor antagonist, suggesting that NK₂ receptor agonists (including BANK) at high doses could produce bronchoconstriction via NK₁ receptor activation (Buckner et al., 1993). We did not pretreat the animals with additional selective NK₁ or NK₂ antagonists during the evaluation of ZD6021. The agonist-mediated bronchoconstriction was markedly displaced to the right in the presence of ZD6021 (by 194- and 77-fold at 10 µmol/kg i.v.) and these changes are consistent with the in vitro potency profile of the compound.

Several studies have reported that additive or synergistic benefits were gained by blocking both the NK₁ and NK₂ receptors. During bronchoconstriction, evoked by antidromic vagal stimulation, either no protection or only partial blockade resulted from selective antagonism of the NK₁ or NK₂ receptor, respectively, whereas cessation of the airway response was brought about by combination treatment with CP 99,994 and SR 48,968 (Savoie et al., 1995). Similarly, a synergistic reduction of airway hyper-responsiveness and inflammatory cell infiltration in ascaris-sensitized monkeys has been reported using these antagonists in combination (Turner et al., 1996). Vagal nerve stimulation of plasma extravasation in the guinea pig lung was mostly prevented by pretreatment with CP 99,994, but leakage was lowest in the presence of both the NK₁ receptor antagonist and the NK₂ receptor antagonist SR 48,968 (Savoie et al., 1995). In the present investigation, low doses of either SR 140333 or ZD7944 (NK₂ receptor antagonist) did not prevent capsaicin-mediated bronchoconstriction. In combination, however, they resulted in a marked rightward displacement of the capsaicin-response curve and this change was similar to that obtained using ZD6021 alone. These data suggest that in the guinea pig and other species, preservation of airway function is best achieved by blockade of both NK₁ and NK₂ receptors during pathophysiological conditions.

The compound MDL 105,212A was the first reported nonpeptide orally available antagonist with high affinity for both NK₁ and NK₂ receptors (Kudlacz et al., 1996). Affinity values (pA₂) for the NK₁ and NK₂ receptors were 8.2 and 8.7, respectively. Capsaicin-induced dyspnea in conscious guinea pigs was blocked after oral administration (ED₅₀ = 50 mg/kg). In our hands, MDL 105,212A yielded pK_B values for the NK₁ and NK₂ receptors of about 7.5 and 6.8 (rabbit pulmonary artery) and 6.9 and 6.4 (human tissue), respectively. In addition, a dose of 19 mg/kg p.o. (or 30 µmol/kg) did not significantly attenuate the fall in airway conductance elicited by administration of ASM-SP or BANK to guinea pigs. By comparison, ZD6021 was a more potent antagonist of these tachykinin-mediated actions in the guinea pig lung. Gerspacher et al. (2000) have recently reported a novel compound, N-[(R,R)-(E)-1-(4-chloro-benzyl)-3-(2-oxo-azepan-3-ylcarbamoyl)-allyl]-N-methyl-3,5-bis-trifluoromethyl-benzamide, that potently blocked agonist-mediated bronchoconstriction with ED₅₀ values of 0.03 and 0.73 mg/kg p.o. for the NK₁ and NK₂ receptors, respectively.

The present study also showed that ZD6021 had good affinity for the human NK₃ receptor, albeit considerably less than that demonstrated for the NK₁ or NK₂ receptors. Nonetheless, this compound potently blocked contraction of the longitudinal muscle of the guinea pig ileum. The pattern of antagonism was unusual; a dose-dependent shift to the right was accompanied by a suppression of the magnitude of contraction even at nanomolar concentrations. These data indicated that ZD6021 demonstrated strong antagonism of the NK₃ receptor, but in a manner distinct from the competitive type observed for the other two tachykinin receptors.

The functional role of the NK₃ receptor in the airways is unclear, although some mention of its biological effects seems warranted, especially with reference to cough. Antitussive effects of NK₂ receptor antagonists have been reported in several studies (for review, see Advenier and Emonds-Alt, 1996). However, the selective NK₃ receptor antagonist SR 142801 was shown to inhibit cough induced by aerosolized citric acid in guinea pigs (Daoui et al., 1998). The site of action for this compound was not likely a direct effect on bronchial smooth muscle (Ellis et al., 1993), rather it is more probable that blockade resulted via neural control of respiratory events. As demonstrated by Bolser et al. (1997), tachykinergic regulation of the cough reflex, stimulated by mechanical irritation of the trachea, occurs within the central nervous system. Although we did not test the effects of ZD6021 in a model of cough, future studies with this compound may provide further clarification of NK₃ receptor involvement in the cough reflex.

The lung is exposed to a number of environmental irritants. In response to these invading particles, unmyelinated C-fibers innervating the lower airways release the tachykinins SP and NKA from their terminal endings, promoting mucous secretion and bronchoconstriction to aid expulsion and to prevent further particle distribution. Moreover, NKB is associated with the cough reflex (Daoui et al., 1998). Alterations of the tachykinergic system result from chronic irritation of the lung, e.g., inhalation of cigarette smoke or from inflammatory diseases, such as asthma. These changes include 1) up-regulation of both NK₁ and NK₂ receptor mRNA in human lungs (Adcock et al., 1993; Bai et al., 1995); 2) decreased levels of neutral endopeptidase, which regulates the availability of the tachykinins (Di Maria et al., 1998); and 3) increased bronchial responsiveness to administration of exogenous SP or NKA (Cheung et al., 1993, 1994). Accordingly, the tachykinins may produce deleterious effects in such pathophysiological conditions and the tachykinergic system may offer an appropriate therapeutic target (Barnes, 1992). Future studies of asthma, cough, or other respiratory (patho)physiology may benefit from tachykinin antagonists, either selective for each of the tachykinin receptors, or those with affinity for multiple tachykinin receptors such as ZD6021.