I[kappa]-B Kinase-2 Inhibitor Blocks Inflammation in Human Airway Smooth Muscle and a Rat Model of Asthma

The nuclear factor-?B (NF-?B) transcription factor plays a key role in the induction of proinflammatory gene expression, leading to the synthesis of cytokines, adhesion molecules, chemokines, growth factors, and enzymes (reviewed recently in Reference 1). These NF-?B-regulated mediators are believed to play a central role in a variety of acute and chronic inflammatory diseases. Increased expression of NF-?B has been demonstrated in the airways of patients with asthma (2-4) and rodent models of asthma (5-8). In addition, other studies have shown that interrupting the actions of NF-?B, using various methodologies, inhibits aspects of the allergic responses in rodent asthma models (9-12). Therefore, it has been suggested that NF-?B plays a pivotal role in asthma pathogenesis and that inhibiting its activity may represent a possible disease-modifying therapy (13).

NF-?B is activated in response to a number of stimuli, including physical and chemical stress, LPS, double-stranded RNA, T- and B-cell mitogens, and proinflammatory cytokines (14-17). NF-KB-induced gene expression is controlled by a complex series of proteins and kinases. In resting cells, the majority of NF-?B is bound to an I-?B inhibitory protein, which holds the complex in the cytoplasm. On appropriate stimulation of the cell, the I-?B protein is phosphorylated and ubiquinated, which leads to subsequent proteasome-mediated degradation. With the I-?B removed, the transcription factor can translocate to the nucleus and activate gene transcription. A critical phosphorylation of the I-?B protein, in the classical pathway, is performed by the I-?B kinase (IKK) complex, which consists of at least three subunits: two catalytic subunits (IKK-1 and IKK-2, also known as IKK-a and IKK-ß) and a regulatory subunit, IKK-? (nuclear factor-?B essential modulator [NEMO]) (18-20). Of the two catalytic subunits, IKK-2 is 20-fold more active than IKK-1 in phosphorylation of I-?B (21), and it has been shown that IKK-2, and not IKK-1, is important in NF-?B activation in vivo (22-26). For this reason, there has recently been a search for a small-molecular-weight inhibitor of IKK-2 for the potential treatment of inflammatory diseases, such as asthma.

The aim of this study was to profile, for the first time, the impact of an IKK-2 inhibitor, TPCA-1 (2-[(aminocarbonyl)amino]-5-[4fluorophenyl]-3-thiophenecarboxamide), on stimulated cultured human airway smooth muscle (HASM) cells and in an in vivo model of antigen-driven airway inflammation. TPCA-1 is a potent and selective IKK-2 inhibitor (negative log of the concentration of drug required to block by 50% [pIC^sub 50^] is 7.74

Some of the results of this study have been reported in the form of an abstract (29)

METHODS

Effect of IKK-2 Inhibitors in Cultured HASM Cells

HASM cells were isolated from lung donor tissue not suitable for transplant and cultured according to the methods detailed in Belvisi and colleagues (30). Ethical approval for the study was obtained together with consent from the relatives. Cells were pretreated with compound (TPCA-1, 10^sup -9^-10^sup -5^ M; SC-514, 10^sup -6^-10^sup -4^ M; budesonide, 10^sup -11^-10^sup -7^ M; or dexamethasone, 10^sup -6^ M) 1 hour before stimulation with IL-1ß (0.1 ng/ml) in the presence of indomethacin (10^sup -5^ M). Parallel experiments with TPCA-1 were performed without indomethacin to determine the impact of cyclooxygenasc (COX) products on the compound activity. In separate experiments, HASM cells were infected with an adenovirus vector containing a dominant negative of either null virus, IKK-1, or IKK-2. Twenty-four hours after stimulation, the supernatants were collected and assayed for cytokine levels (granulocyte-macrophage colony-stimulating factor [GM-CSF], granulocyte [G]-CSF, and IL-8) and cell viability.

To demonstrate that treatment with TPCA-1 blocked NF-?B nuclear translocation, similar experiments were performed as outlined above, except the nuclear fractions were collected 1 hour after stimulation and NF-?B DNA binding was assessed using an electro mobility shift assay (EMSA). In addition, the functional effect of TPCA-1 on NF-?B translocation was measured using an NF-?B-luciferase reporter assay.

To show that the effect observed with TPCA-1 was not specific to IL-1ß-induced cytokine production, parallel experiments were performed on TNF-a (10 ng/ml)-induced GM-CSF and IL-8 release, IL-13 (1 ng/ml)-induced eotaxin release, and serum (3% fetal calf serum)-induced cell proliferation.

Effect of TPCA-1 in an In Vivo Model of Antigen-induced Airway Inflammation

To aid determination of appropriate dose levels and times for the main study, a small experiment was performed to determine systemic drug levels after oral administration. In the main study, sensitized male Brown Norway rats were challenged with aerosolized saline or ovalbumin (10 g/L) for 30 minutes. There were two protocols used for assessment of the impact of TPCA-1 on airway inflammation: (1) analysis of samples taken 6 hours after antigen challenge to determine effect of the compound on inflammatory gene and protein expression and (2) analysis of samples taken 24 hours after antigen challenge to determine effect of the compound on the cellular influx. These time points were chosen from previous data generated by our group (31).

A small satellite study, based on the first protocol, was performed alongside the main study to assess the effect of the compound on NF-?B pathway activation by using a Trans-AM p65:DNA binding plate assay (Active Motif, Belgium).

Assessment of the effect of the compound on inflammatory cytokine gene and protein expression (IL-4, IL-5, IL-13, IL-1ß, TNF-a, and eotaxin) was performed on a Taqman real-time polymerase chain reaction machine (ABI PRISM 7000; Applied Biosystems, Foster City, CA) and by ELISA, respectively. The effect of the compound on airway inflammatory cell influx 24 hours after antigen challenge was determined in the bronchoalveolar lavage fluid and lung tissue.

In addition, a parallel study was performed using the same sensitizing and challenge protocol in which the effect of TPCA-1 (30 mg/kg) on the late asthmatic response and NF-?B nuclear translocation was measured.

Full details of the methods outlined above are contained in an online supplement. Specifics on data and statistical analysis are presented in the figure legends.

RESULTS

Effect of the IKK-2 Inhibitors on Stimulated Cultured HASM Cells

Stimulation of cultured HASM cells with IL-1ß caused an increase in GM-CSF (60 ± 20 increased to 2,900 ± 900 pg/ml) and G-CSF (5 ± 3 increased to 1,730 ± 480 pg/ml) production; these increases were inhibited by TPCA-1 (concentration required to elicit 50% of the maximal effect [EC^sub 50^]), 6.4 ± 0.1, and maximal effect [E^sub max^], 97 ± 1%; and EC^sub 50^, 6.8 ± 0.1, and E^sub max^, 99 ± 1%, respectively; Figure 1). Interestingly, although both dexamethasone (E^sub max^ on average, ~ 95%) and budesonide (EC^sub 50^, 9.6 ± 0.1, and E^sub max^, 93 ± 4%) inhibited GM-CSF production, neither impacted on the release of G-CSF to the same extent as TPCA-1 (Emax on average, ~ 30-40%; EC^sub 50^, 9.1 ± 0.1, and E^sub max^, 44 ± 9%, respectively; Figure 1). IL-1ß-stimulated HASM cells released IL-8 (400 ± 120 increased to 32,770 ± 6,790 pg/ml), the production of which was also inhibited by TPCA-1 (EC^sub 50^, 6.8 ± 0.2, and E^sub max^, 99 ± 1%; Figure 2). Similar to the findings with G-CSF, neither dexamethasone nor budesonide impacted on the release of IL-8 to the same extent as TPCA-1 (E^sub max^ on average, ~ 70-80%; EC^sub 50^, 9.1 ± 0.1, and E^sub max^, 78 ± 4%, respectively; Figure 2). In the parallel experiments without indomethacin, the levels of IL-1ß induced the following: GM-CSF release was reduced (1,078 ± 440 pg/ml), G-CSF release was increased (5,775 ± 995 pg/ml), and IL-8 remained the same (34,980 ± 8,870 pg/ml). The presence of COX products did not alter the potency or efficacy of TPCA-1 (EC^sub 50^, 6.4 ± 0.2, and E^sub max^, 100 ±0%; EC^sub 50^, 6.7 ± 0.1, and E^sub max^, 99 ± 0%; EC^sub 50^, 6.7 ± 0.3, and E^sub max^ 98 ± 1%, for GM-CSF, G-CSF, and IL-8, respectively). Although the potency values for budesonide obtained in the assays without indomethacin remained similar to the results with COX inhibition, the efficacy values were altered (EC^sub 50^, 9.3 ± 0.1, and E^sub max^, 82 ± 3%; EC^sub 50^, 9.5 ± 0.1, and E^sub max^, 71 ± 2%; EC^sub 50^, 9.2 ± 0.2, and E^sub max^, 64 ±5%, for GM-CSF, G-CSF, and IL-8, respectively). To provide further confirmation of the effect observed with TPCA-1, similar data were obtained using a structurally different IKK-2 inhibitor, SC-514, and dominant negative IKK-2, and not IKK-1, technology (Figure 2). Inhibition of GM-CSF and G-CSF, similar to that seen with TPCA-1, was also observed with SC-514 and dominant negative IKK-2 infected cells (data not shown). Although the data for SC-514 demonstrate that a high concentration is needed to inhibit mediator release, this result is not unexpected, because the activity (IC^sub 50^) on the isolated kinase is 11.2 ± 4.7 µM (28). Because of the limited supply of dominant negative virus, we were only able to perform these experiments in cells obtained from two different donors. The low numbers made statistical analysis impossible; however, the clear impact of dominant negative IKK-2, and not null virus (suggesting that any activation of Toll-like receptors by the virus does not impact on the assay system used) or IKK-1, on cytokine release is very compelling and further complements and supports the results with the two pharmacologic inhibitors.

To demonstrate that TPCA-1 blocks the nuclear translocation of NF-?B, we performed a series of experiments. The EMSA performed with the nuclear extracts shows that TPCA-1 caused a concentration-related inhibition of NF-?B nuclear translocation (Figure 3A). To more accurately quantify the levels of NF-?B in the nuclear fraction, we first determined which rel family members made up the IL-1ß-induced EMSA signal. The supershift antibodies raised against p50 and p65 both shifted the EMSA band, whereas the other rel family member antibodies did not (Figure 3B). This result suggested that the NF-?B present in the nucleus, in this assay system, at this time, consisted of p50:p65 heterodimers. We went on to use a more quantitative, commercially available assay to demonstrate that TPCA-1, and not dexamethasone, caused a significant concentration-related inhibition of p65 nuclear translocation (Figure 3C). We confirm these results functionally using an NF-?B reporter assay (Figure 3D).

To demonstrate that the impact of TPCA-I was not specific to IL-1ß stimulation and can reduce a function of the HASM cells, proliferation, parallel studies were performed using TNF-a, IL-13, or fetal calf serum as a stimulant. TNF-a stimulation caused a significant increase in GM-CSF and IL-8 release (13 ± 4 to 390 ± 70, and 90 ± 20 to 4,330 ± 560 pg/ml, respectively; Figure 4). No increase in G-CSF release was measured (data not shown). TPCA-I treatment caused significant inhibition of both cytokines with potency and efficacy values similar to the results with IL-1ß stimulation (EC^sub 50^, 6.3 ± 0.5, and E^sub max^, 91 ± 2%; and EC^sub 50^, 6.9 ± 0.2, and E^sub max^, 99 ± 1%, respectively; Figure 4). IL-13 stimulation caused an increase in eotaxin release (620 ± 80 to 1,340 ± 80 pg/ml), which was significantly inhibited by TPCA-I (EC^sub 50^, 6.2 ± 0.2; Figure 4). Fetal calf serum-induced DNA synthesis was significantly inhibited by TPCA-1, with an EC^sub 50^ value of 6.0 ± 0.3 (Figure 4)

There was no significant impact on cell viability, as measured by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assays, with any of the treatments (MTT values for TPCA-1 at the top concentration tested, 10 µM, were 0.7 ± 0.1, 0.7 ± 0.1, 0.8 ± 0.1, and 0.6 ± 0.1 for the experiments with IL-1ß [with indomethacin], IL-1ß [without indomcthacin], TNF-a, and IL-13, respectively, compared with 0.8 ± 0.1, 0.8 ± 0.1, 0.9 ± 0.1, and 0.6 ± 0.1 for the stimulated control).

Effect of the IKK-2 Inhibitor on Antigen-induced Airway Inflammation

Analysis of systemic compound levels after oral administration demonstrated that TPCA-1 is orally bioavailable. It appeared that the oral bioavailability was approximately linear, in that the more compound administered, the greater the systemic exposure (Figure 5). Administration of 30 mg/kg resulted in plasma compound levels of approximately 4 µg/ml, which were maintained for at least 4 hours. This concentration of compound equates to in excess of 10 µM, which in the HASM cell experiments was sufficient to cause maximal inhibition. Although direct extrapolations like this are in no way conclusive, because of unknown parameters, such as plasma binding and target tissue penetration, we used these data to select 30 mg/kg as the top dose used and for the times of drug administration in subsequent studies.

In the small satellite study, lung tissue levels of p65 not bound to I-?B were increased 6 hours after antigen challenge, indicating an increase in NF-?B pathway activity. Administration of the IKK-2 inhibitor caused a dose-related reduction in p65 levels, suggesting an inhibition of NF-?B pathway activation via blocking the activity of IKK-2 (Figure 6).

Antigen challenge caused a significant increase in lung tissue cytokine gene expression. Changes in threshold cycle (ct) values (the lower the number, the more target gene present) for IL-1ß, IL-4, IL-5, IL-13, eotaxin, and TNF-a in the saline/vehicle and ovalbumin/vehicle groups were as follows: 17.1 ± 0.2 and 13.4 ± 0.5, 25.3 ± 0.3 and 22.9 ± 0.6, 25.5 ± 0.3 and 21.1 ± 0.6, 25.5 ± 0.4 and 21.4 ± 0.5, 15.5 ± 0.2 and 12.1 ± 0.3, and 17.2 ± 0.1 and 13.9 ± 0.7, respectively; Figure 7). This increase in gene expression was significantly inhibited by TPCA-1 in a dose-related manner (Figure 7). Interestingly, the profile of gene expression inhibition seen with administration of TPCA-1 was comparable to the impact on NF-?B pathway activation, suggesting a causative affect. Not all Th-2 cytokine genes are believed to contain a NF-?B binding site in their promoter, which would suggest that the effect of the compound on the expression of these genes is occurring through an indirect mechanism, such as impacting on other cytokines or the recruitment/activation of the inflammatory cells.

The glucocorticoid comparator budesonide significantly inhibited antigen-induced cytokine gene expression.

The antigen-induced increase in lung tissue cytokine gene expression was associated with a significant increase in IL-1ß and eotaxin protein expression (Figure 8). At the time point of tissue sampling chosen in this study, however, there was not an antigen-induced increase in TNF-a, IL-4, or IL-13 protein expression. The IKK-2 inhibitor caused a dose-related significant inhibition of antigen-induced IL-1ß and eotaxin protein expression; in addition, basal levels of these cytokines were also significantly reduced (Figure 8). The impact of TPCA-1 on levels of these cytokines, and on lung tissue eosinophilia (Figure 9), in the control-challenged group would imply that a portion of these indices are under the control of IKK-2 activity in the saline-challenged lung. However, it could suggest that because there appeared to be no effect on p65:DNA binding in this group, the compound is having off-target effects. The glucocorticoid comparator budesonide significantly inhibited antigen-induced IL-1ß and eotaxin protein expression.

Twenty-four hours after antigen challenge, there was a significant increase in bronchoalveolar lavage eosinophil and neutrophil numbers and in lung tissue eosinophilia (Figure 9). TPCA-1 caused a significant dose-related inhibition of these inflammatory cells to a similar extent as that produced by budesonide (Figure 9). Antigen challenge or the compounds had no effect on the number of other white cells measured (data not shown). Arbitrary scoring of histologie sections taken from the fixed and processed lung tissue resulted in a similar profile of eosinophilia as that obtained with the cell digest and no increase in neutrophil number (Figure 9).

Using this antigen-driven model, our group has shown that an "early asthmatic response," which consists of acute, steroid-insensitive bronchospasm, occurs immediately after antigen challenge (32). After approximately 10 minutes, this bronchospasm dissipates, and then approximately 1 to 3 hours later, a "late asthmatic response" is observed. Using this experimental protocol, the early asthmatic response probably occurs during the 30 minutes of antigen challenge. The antigen challenge in this experiment caused a robust late asthmatic response, which was inhibited by treatment with TPCA-1 and budesonide (Figure 10). To confirm that TPCA-1 had prevented NF-?B nuclear translocation in this preclinical asthma model, EMSA analysis was performed using nuclear extracts from lung tissue. The autoradiograph shown in Figure 10 clearly shows that TPCA-1, and not budesonide, has blocked the antigen-induced NF-?B nuclear translocation.

DISCUSSION

Until recently, the contractile property of airway smooth muscle was regarded as its sole contribution to the pathogenesis of asthma. Indeed, early studies on HASM function concentrated on the contractile aspects to explain airway hyperreactivity, a characteristic of asthma in which the airways show an exaggerated response to a large number of different stimuli acting through many pathways. This view, however, is rapidly changing because airway smooth muscle can undergo hyperplasia and/or hypertrophy, which can lead to structural changes in the airway wall, and may contribute to the development of persistent airway obstruction and increased nonspecific airway hyperreactivity in chronic severe asthma (33). In addition, more recent research has concentrated on the secretory capacity of airway smooth muscle. The HASM cell is one of the more prominent structural cells in the airway and is a rich source of many cytokines, chemokines, and inflammatory mediators, which may modulate airway inflammation in asthma (34). In this study, TPCA-1 inhibited the release of all IL-1ß-induced cytokines measured that were released from the stimulated HASM cells. Similar data were obtained with SC-514, another IKK-2 inhibitor, and cells infected with a virus containing dominant negative IKK-2, which further strengthens the results observed with TPCA-1. The majority of the studies with HASM cells were performed in the presence of indomethacin because COX products have been shown to alter the expression of GM-CSF and G-CSF in HASM cells (35-37). However, because NF-?B has been linked to the transcriptional control of the inducible form of COX, COX-2 (38, 39), it was decided to determine the effect of TPCA-1 in the same assay system without indomethacin. The results showed that the presence of COX products did not impact on the potency or efficacy of the IKK-2 inhibitor, even though, without indomethacin, we observed an expected decrease in GM-CSF release and increase in G-CSF release. Interestingly, although the potency of budesonide was not altered, the efficacy on G-CSF release increased. It seems likely that this additional efficacy could be due to the effect of the steroid on COX-2 formation, which would indirectly reduce the level of G-CSF release.

The series of studies to confirm that TPCA-1 was blocking NF-?B translocation demonstrated very clearly that, in this cell system, at 1 hour after stimulation with IL-1ß, the NF-?B translocated to the nucleus were heterodimers of p50:p65. The NF-?B EMSA, the p65 plate assay, and the functional NF-?B reporter assay results showed that TPCA-1 stopped the translocation of these heterodimers. This suggests that the compound is inhibiting the removal of I-?B, thus stopping transcription. However, IKK1/2 has also been suggested to be one of the many kinases to directly phosphorylate members of the NF-?B rel family, and other components of the transcriptional machinery, an effect which in turn is likely to impact on the transcriptional responses (40). These phosphorylation events were not studied in this current article and therefore we cannot rule out if any effect(s) of the compound are due to other post-I-?B regulation of NF-?B pathways.

To demonstrate that the impact of TPCA-1 was not specific to IL-1ß stimulation and can reduce functional responses of the HASM cells, such as proliferation, parallel studies were performed using TNF-a, IL-13, or fetal calf serum as stimulants. These results showed that TPCA-1 can impact on TNF-a-induced GM-CSF and IL-8 release, IL-13-induced eotaxin release, and DNA synthesis, which is a marker of cell proliferation.

The results in their entirety would suggest that an IKK-2 inhibitor would impact on airway smooth muscle-driven disease pathogenesis. Interestingly, although GM-CSF release was inhibited by treatment with both of the glucocorticoids tested, release of G-CSF and IL-8 were only partially impacted. Similar limited impact of steroid treatment on G-CSF and IL-8 release from HASM cells has been published (41, 42). Although it is not known why these cytokines appear to be less sensitive to glucocorticoid treatment, it is of note that these two cytokines are involved in the growth, chemoattraction, and maintenance of neutrophils, cells believed to be involved in the pathogenesis of steroid-insensitive diseases, such as severe asthma and chronic obstructive pulmonary disease.

The IKK-2 inhibitor reduced antigen-induced allergic airway inflammation in sensitized Brown Norway rats. Models of allergic pulmonary inflammation in the rat and mouse are widely used to assess the impact of potential new therapies for asthma. Many of the inflammatory features seen in the asthmatic airway are present in these models, including increased cytokine gene and protein expression (which includes the Th2-type cytokines), in lung tissue and airway eosinophilia (43-45). In this study, oral administration of TPCA-1 (30 mg/kg) resulted in plasma concentration levels in excess of 10 µM for up to 4 hours. This was the concentration required in the human cell-based assay system used to achieve maximal inhibition. This level of systemic exposure reduced the antigen-induced increase in p65 DNA binding and nuclear translocation of NF-?B, which suggests that the target kinase, IKK-2, had indeed been inhibited. The result from the smaller satellite study not only acted as a proof of concept but also mirrored the effect of the compound on gene and protein cytokine expression, suggesting a causative link. Indeed, the magnitude of inhibition of cytokine production by the IKK-2 inhibitor was similar to that achieved with a high dose of a clinically relevant glucocorticoid, budesonide. The reduction in antigen-induced cytokine production by TPCA-1 was associated with a similar inhibition of airway eosinophilia and neutrophil numbers, again suggesting a functional link. In addition, the late asthmatic response observed in this model was completely blocked by the top dose of TPCA-1, indicating inhibition of IKK-2 may be of clinical benefit by alleviating this symptom of asthma.

In conclusion, this study has shown that an IKK-2 inhibitor is capable of inhibiting an inflammatory response in HASM cells and allergic airway inflammation in the rat. The anti-inflammatory profile of the IKK-2 inhibitor, described in this study, would suggest that compounds of this class could, in certain respects, possess a beneficial efficacy profile compared with that of a steroid. This study represents the first demonstration of anti-inflammatory activity in the airways using a small-molecule-inhibitor approach, suggesting the utility of IKK-2 as a disease-relevant therapeutic target.

Conflict of Interest Statement: M.A.B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. E.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. S.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. M.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.D.A. was employed as a contractor from February 2003 to October 2004 by GlaxoSmithKline (GSK; scientific recruitment group agency). Currently, he does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. R.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. S.H.-Y. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.T.P. is a full-time employee of GSK PLC. C.J.W. is a full-time employee of GSK PLC. R.J.S. is an employee of GSK PLC. T.J.S. has been employed by GSK or its legacy companies from 1965 until the present time. Any stockholdings are a result of Employee Schemes run by GSK, which employees can choose to participate in. M.G.B. has been reimbursed by GSK, Altana, and Pfizer, and for attending conferences/meetings. She has served as a consultant to GSK and Novartis and has received research grants from GSK and Novartis. Between 1997 and 2000, she was an employee of Aventis Pharma, and has served on advisory boards for Aventis Pharma, Altana, and GSK.

Acknowledgment: The authors thank GlaxoSmithKline for the generous donation of the TPCA-1 compound and acknowledge the chemists involved in the synthesis: Stefano Livia, James Callahan, Yue Li, and Robert Blade. In addition, they thank Professor Michael Karin for generously providing the IKKI/2 dominant negatives and Miss H. Keen for help with the preparation of this manuscript.

© 2005 American Thoracic Society Provided by ProQuest LLC. All Rights Reserved.