Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Airway response to antigen challenge in asthmatic individuals has been divided into early and late phases based upon transient declines in lung function that occur immediately or 6 to 8 h following exposure, respectively (Hersheimer, 1952). The late phase response is characterized by airway hyperresponsiveness and influx of inflammatory cells including T cells, eosinophils, neutrophils, and monocytes (Metzger et al., 1986; Aalbers et al., 1993). In fact, the increase in eosinophil numbers and activation state are distinguishing features of the disease, making their selective elimination a target for the treatment of asthma.

One approach to inhibit infiltration of eosinophils into tissues is by interfering with cell adhesion molecules that facilitate the rolling and firm adhesion of these cells prior to diapedesis (Wardlaw, 1999). Firm adhesion occurs via interactions between cell adhesion molecules such as LFA-1 and Mac-1 with their endothelial counterpart ICAM-1 or 41 with VCAM-1. Both ICAM-1 and VCAM-1 can be up-regulated in the airways following antigen challenge (Bentley et al., 1993; Ohkawara et al., 1995). Some selectivity in the cells recruited in response to antigen may occur since 41 expression appears to be primarily restricted to eosinophils, lymphocytes, and monocytes, with variable expression on neutrophils (Kirveskari et al., 2000; Ibbotson et al., 2001). The "sparing" of neutrophils from the potential effects of 41 inhibition has made it a more attractive target than inhibition of 2 integrin interactions, e.g., LFA-1/ICAM-1.

A critical role for 41 in tissue eosinophil infiltration has been established using antibodies raised against the adhesion molecule in models of allergic pulmonary inflammation in diverse species including mice (Nakajima et al., 1994; Chin et al., 1997; Kanehiro et al., 2000), rats (Richards et al., 1996; Schneider et al., 1999), guinea pigs (Pretolani et al., 1994; Das et al., 1995), and sheep (Abraham et al., 1994). There are a limited number of publications that describe efficacy for small molecule antagonists of 41 interactions with VCAM-1 and fibronectin. The majority of peptide-based small molecule antagonists have been developed using the observation that Leu-Asp-Val is the minimal binding sequence of the CS-1 region on fibronectin, which overlaps with the binding site for 41 on VCAM-1 (Pulido et al., 1991). Two small molecule 41 antagonists have been evaluated in the allergic sheep model following aerosol administration. Efficacy was reported for both the CS-1 ligand mimic phenylacetyl-L-leucyl-L-aspartyl-L-phenylalanyl-D-prolineamide (Abraham et al., 1997) and BIO1211 (Abraham et al., 2000) as measured by modification of antigen-induced alterations in cell infiltration and pulmonary function.

In this study, we compare the in vitro and in vivo efficacy for several small molecule 41 antagonists. CP-664511 was identified as an inhibitor with high potency based on its ability to inhibit Jurkat cell binding to VCAM-1 and to attenuate antigen-induced bronchoalveolar lavage (BAL) eosinophil infiltration in a murine model of pulmonary inflammation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

In Vitro Inhibition of Jurkat Cell Binding to VCAM-1. Jurkat cells were activated upon incubation with RPMI 1640 medium containing 1 mM MnCl₂ for 30 min at 37°C. Activated cells were washed twice with serum-free medium containing media and fluorescently labeled by incubation with calcein AM (Molecular Probes, Eugene, OR) for 30 min at 37°C. Cells (140,000 per well) were aliquoted into a 96-well polypropylene plate (Nalge Nunc, Naperville, IL) in PBS containing 1 mM calcium, 1 mM magnesium, and 1 mg/ml bovine serum albumin (PBSB). Compounds were added to wells and incubated for 45 min at 37°C. An aliquot of 150 µl of cell suspension (=100,000 cells) plus compounds were transferred to soluble VCAM-coated plates and incubated for 45 min at 37°C. Nonadherent cells were removed by shock dumping and washing twice in PBSB. Finally, 200 µl of PBSB were added to adherent cells and the percentage of control fluorescence was determined in a CytoFluor II fluorescent plate reader. IC₅₀ values were determined from a dose-response curve run in triplicate. In those experiments examining the effects of protein binding on the IC₅₀, PBSB was replaced with 100% serum.

Mouse Model of Allergic Pulmonary Infiltration. All procedures were approved by the Pfizer Animal Care and Use Committee. Female Balb/c mice (4-6 weeks of age) were injected i.p. with chick egg ovalbumin (100 µg) adsorbed to alum (4.5 mg) (Pierce Chemical, Rockford, IL) in a volume of 200 µl on days 0 and 7. On days 21 and 22, mice were briefly anesthetized with metofane to facilitate intratracheal instillation made in a volume of 50 µl. The vehicle for intratracheal instillation was a mixture of 25% ethanol/40% polyethylene glycol 200/35% water. Mice were exposed 45 min later to aerosols of PBS or ovalbumin (3%) delivered for 30 min by a DeVilbiss ultrasonic nebulizer (Somerset, PA) into an 18 × 43-cm Plexiglas chamber. The effects of anti-4 antibody administered into the airways were evaluated following intratracheal instillation of PS/2 (American Type Culture Collection, Manassas, VA; CRL 1911, 38 µg daily) as described. To evaluate effects of anti-4 antibody following systemic administration, PS/2 was administered by intraperitoneal injection 1 h before the first (400 µg) and second (100 µg) antigen challenges or 3, 24, and 48 h after the second antigen challenge (400, 100, and 100 µg, respectively). In those studies in which the effects of CP-664511 following systemic administration were evaluated, the compound was dosed by s.c. injection twice daily before and after antigen challenge.

Reagents. Unless otherwise specified, compounds were obtained from Sigma-Aldrich (St. Louis, MO). All 41 antagonists were synthesized at Pfizer Central Research as described elsewhere (Duplantier et al., 2001; Fig. 1).

Statistics. Data are expressed as mean ± S.E.M. Analyses were performed using two-way analysis of variance followed by multiple comparisons procedures using Tukey's test. When the normality test failed, analysis of variance on ranks was used and multiple comparisons made with Dunn's method. Significance was assigned at p < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Inhibition of Jurkat Cell Binding to VCAM-1 in Vitro. All compounds dose dependently inhibited Jurkat cell binding to VCAM-1 with potencies (IC₅₀) ranging from 0.5 nM (CP-664511) to 39 nM (CP-609643) (Table 1). However, the potency of all compounds was reduced when binding was evaluated in the presence of 100% serum. The magnitude of potency loss varied with CP-609643 being the least (3-fold) and BIO1211 the most (42-fold) affected. CP-664511 remained the most potent compound even in the presence of serum (IC₅₀ = 5 nM).

Effects of Local Airway Instillation of 41 Antagonists on Antigen-Induced Pulmonary Eosinophil Infiltration and Inflammatory Mediator Release in Vivo. In ovalbumin-sensitized mice, two daily aerosol exposures increased total BAL white blood cells from 4.16 ± 0.46 e4 cells/ml (OVA-sensitized, PBS aerosol-challenged) to 18.4 ± 1.66 e4 cells/ml (OVA-sensitized, OVA aerosol-challenged) 72 h following the second challenge. Eosinophils comprised 37% of the total white blood cells following OVA exposure. BAL lymphocytes and neutrophils were minimal (1-2%) in both PBS and OVA aerosol-exposed animals and hence were not considered further.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Chemical structures for BIO1211, CP-664511, CP-609643, CP-619700, and TR-14035.

View this table: [in this window] [in a new window]   TABLE 2 Effects of 41 antagonism on antigen-induced inflammatory mediator release in BALF Data are mean ± S.E.M. from nine to ten animals per group. Vehicle or CP-609643 (200 µg) was administered by intratracheal instillation 45 min before aerosol challenge. TNF- levels were evaluated in lavage fluids 30 min after saline or ovalbumin 3% aerosols as described under Materials and Methods. Eotaxin and IL-4 levels were evaluated in the same BALF samples 24 h after aerosol exposure.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Effects of 41 antagonists on BALF eosinophil influx in response to antigen challenge in allergic mouse model. Data are mean ± S.E.M. from n = 10 to 20 mice per dose per treatment group and are expressed as percentage of inhibition of BAL eosinophil numbers in vehicle-treated animals. Mice received compounds by intratracheal instillation in a volume of 50 µl 45 min before antigen challenge. Dosing and exposure were repeated on the following day. , significant (p < 0.05) difference from vehicle response.

Effects of Systemic VLA-4 Antagonism on Antigen-Induced Eosinophil Infiltration in Vivo. Intraperitoneal administration of the anti-4 PS/2 to OVA-sensitized mice either before or after antigen challenge inhibited BAL eosinophil infiltration by ~60% (Fig. 3). CP-664511 dose dependently (1-10 mg/kg, s.c.) reduced antigen-induced eosinophil infiltration when administered by the subcutaneous route twice daily both before and after antigen challenge (Fig. 4). Inhibition was significant only at the 10 mg/kg dose at which ~58% inhibition of eosinophil infiltration was observed.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Comparison between inhibition of BALF eosinophil influx in response to steroids or anti-4 antibody. Data are mean ± S.E.M. from n = 10 to 20 mice per treatment group and expressed as percentage of inhibition relative to appropriate vehicle control. Anti-4 antibody was either administered by i.t. instillation (76 µg) or i.p. injection (500 µg) 1 h before or 3 h after aerosol challenge. , significant difference from vehicle response (p < 0.05).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Effects of CP-664511 on antigen-induced BALF eosinophil influx. Data are mean ± S.E.M. from 10 to 20 mice per treatment group. Animals received s.c. injections of CP-664511 1 h before and 5 h after aerosol challenge on 2 consecutive days. Compound administration was then continued twice daily for 2 additional days until BAL was performed. , significant difference from vehicle-treated animals (p < 0.05).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Neutralizing antibodies against 4 have been shown to ameliorate various aspects of late phase allergic airway responses such as inflammatory cell infiltration and hyperresponsiveness (Foster, 1996; Elices, 1999). These effects have made inhibition of 41/VCAM-1 interactions an attractive target for the development of small molecule antagonists. Our studies demonstrate anti-inflammatory activity for several peptidomimetic inhibitors developed through modification of the Leu-Asp-Val sequence, which is a critical region for the binding of 4 to the CS-1 region on fibronectin (Duplantier et al., 2001). Compounds resulting from these efforts include CP-664511, which is almost 10 times more potent than BIO1211 in inhibiting Jurkat cell binding to VCAM-1 in the absence of serum and nearly 30 times more potent in its presence. The in vitro activity obtained in these experiments for BIO1211 (IC₅₀ = 4 nM) is comparable with data obtained by others (IC₅₀ = 4 nM) (Lin et al., 1999). The rank order of compound potency in the Jurkat cell binding assay was generally consistent with in vivo activity as reflected by inhibition of pulmonary eosinophil infiltration in response to aerosolized antigen challenge. BIO1211, although equieffective in vitro with other compounds such as CP-619700, failed to demonstrate significant inhibition of eosinophil infiltration even at doses up to 10 mg/kg, i.t. and suggests that factors other than in vitro activity contribute to in vivo efficacy in this model. In contrast, BIO1211 was efficacious when delivered by aerosol or intravenous routes in an allergic sheep model of asthma using inflammatory cell influx and airway hyperresponsiveness as endpoints (Abraham et al., 2000).

The maximum inhibition of BAL eosinophil infiltration produced by an antibody to 4 or small molecule 41 antagonist in this murine model of allergen-induced pulmonary inflammation was found to be ~60%. In contrast, beclomethasone administered under the same conditions was capable of producing nearly complete attenuation of the antigen-induced inflammatory cell influx. One explanation for the submaximal effect of 41 antagonists may be the functional redundancy that exists between adhesion molecules, in particular, 1 and 2 integrins. Firm adhesion of eosinophils to vascular endothelial cells and their transendothelial migration are also mediated via LFA-1 and ICAM-1 interactions (Wardlaw, 1999). ICAM-1, like VCAM-1, is up-regulated at sites of allergic inflammation in conditions such as asthma and atopic dermatitis (Kyan-Aung et al., 1991; Montefort et al., 1992). Monoclonal antibodies to CD18 and 4 reduced BAL eosinophilia by 62 and 58%, respectively, in a guinea pig model of Sephadex-induced pulmonary inflammation (Das et al., 1995). When the antibodies were coadministered, total inhibition of BAL eosinophilia was observed. In a rat model of allergic pulmonary inflammation, treatment with anti-4 antibody reduced BAL eosinophilia (66%), which was similar to efficacy for anti-2 integrin antibody (54%) (Schneider et al., 1999). Again, a combination of the two antibodies produced more complete inhibition (98%) than either alone.

4 binds not only to VCAM-1, but also to alternatively spliced CS-1 fibronectin on the extracellular matrix. This interaction has been proposed to result in cell activation, such as the mast cell (Ra et al., 1994), and may explain the efficacy of 41 antagonism in early phase response to antigen challenge as has been reported in rats (Hojo et al., 1998) and sheep (Abraham et al., 1997, 2000). In our studies, intratracheal instillation of CP-609643 at a dose that significantly attenuated eosinophil infiltration failed to affect levels of the mediators known to be important in the recruitment of these cells, including TNF-, eotaxin, and IL-4. Both IL-4 and TNF- can be released by murine mast cells following IgE cross-linking (Gordon and Galli, 1991; Yamaguchi et al., 1997), although other sources include T cells and macrophages, respectively. Importantly, both of these cytokines have been shown to up-regulate VCAM-1 in vitro and in vivo (Iademarco et al., 1995; Sanz et al., 1997; Hickey et al., 1999).

Delivery of 41 antagonists directly to the airways may offer advantages in terms of efficacy and safety over systemic administration. Although this may be appropriate for attenuation of certain end points, such as airway hyperresponsiveness, it is not clear that local inhibition of 41 interactions affects inflammatory cell influx. In our studies, intratracheal instillation of PS/2 antibody (total dose = 76 µg) against 4 before antigen challenge modestly (~40%) reduced BAL eosinophilia. Systemic administration of the same antibody did, however, significantly reduce cell influx regardless of whether PS/2 was administered before (62% inhibition) or after (66% inhibition) allergen challenge. These data are consistent with those of Abraham et al. (1994) who found that intravenous but not aerosol administration of the anti-4 monoclonal antibody HP 1/2 reduced airway inflammation in the allergic sheep model. However, Henderson et al. (1997) reported intranasal anti-4 antibody treatment (doses of 11-56 µg administered before antigen challenge) attenuated eosinophil infiltration in a mouse model of allergic pulmonary inflammation to a similar extent as that following i.p. injection. Small molecule 41 antagonists administered by intratracheal instillation are well absorbed by the lung, exhibiting good systemic exposure and extended half-life, which complicates interpretation of the site of their anti-inflammatory action (Sargent et al., 2000; unpublished data). It is interesting to note, however, that the dose of CP-664511 required to inhibit eosinophil infiltration ~60% following s.c. administration is ~10 times higher than the dose required to produce the same effect following intratracheal instillation.

Eosinophils also express 47, which may contribute to their function (Wardlaw, 1999). It is unlikely that this ligand plays an important role in eosinophil infiltration since 41 antagonists including BIO1211 and CP-664511 are not potent inhibitors of 47/mucosal addressin CAM interactions and the magnitude of inhibition of BAL eosinophilia that they produce (60%) is consistent with that produced by anti-4 antibodies (60%). Furthermore, the efficacy of TR-14035 in vivo is not improved relative to other compounds evaluated despite the fact that it has been reported to inhibit 47 with a potency approximately 10 times greater than its effects on 41 (Sircar et al., 1999). Finally, there were no differences between 7-deficient and wild-type mice in inflammatory cell influx into airways of mice receiving Aspergillus extract (Venkayya et al., 1998). Although these data do not support a role for 47 in antigen-induced pulmonary eosinophil infiltration, in vitro studies have suggested that the adhesion molecule may be important for eosinophil survival (Meerschaert et al., 1999).

In conclusion, these studies demonstrate the ability of small molecule 41 antagonists to inhibit airway eosinophil infiltration in a murine model of allergic pulmonary inflammation. CP-664511 was found to be the most potent agent evaluated both in vitro and in vivo. These agents may be useful in the treatment of allergic pulmonary inflammation and are currently undergoing clinical trials in asthmatic patients (Bolger, 2000; Sargent et al., 2000).