Published ahead of print on July 20, 2006, doi:10.1164/rccm.200601-048OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200601-048OC
Whole-Body Periodic Acceleration Modifies Experimental Asthma in SheepDivision of Pulmonary Disease and Critical Care Medicine, Miller School of Medicine, University of Miami at Mount Sinai Medical Center; and Department of Neonatology, Mount Sinai Medical Center, Miami Beach, Florida Correspondence and requests for reprints should be addressed to William M. Abraham, Ph.D., Department of Research, Mount Sinai Medical Center, 4300 Alton Road, Miami Beach, FL 33140. E-mail: abraham{at}msmc.com
Rationale: Nitric oxide is released from vascular endothelium in response to increased pulsatile shear stress. Nitric oxide inhibits mast cell activation and is antiinflammatory and therefore might be protective in asthma. Objectives: We determined if a noninvasive motion platform that imparts periodic sinusoidal inertial forces to the whole body along the spinal axis (pGz) causing release of endothelial nitric oxide modulates experimental asthma in sheep.
Methods: Allergic sheep were untreated (control) or were treated with pGz alone or after receiving intravenously the nitric oxide synthase inhibitor Nw-nitro-L-arginine methyl ester (L-NAME) before aerosol challenge with Ascaris suum, and the effect on antigen-induced airway responses was determined. Bronchoalveolar lavage cells obtained 6 h after antigen challenge were analyzed for nuclear factor-
Results: pGz treatment for 1 h before antigen challenge reduced the early airway response and blocked the late airway response but did not prevent the antigen-induced airway hyperresponsiveness 24 h after challenge. Administration of L-NAME before pGz completely reversed this protection, whereas L-NAME alone did not affect the antigen-induced responses. NF- Conclusion: Whole-body pGz modulates allergen-induced airway responses in allergic sheep.
Key Words: animal models asthma therapy nitric oxide noninvasive periodic acceleration Nitric oxide (NO) has varied physiologic and pathophysiologic roles in biological systems. NO is formed during the conversion of L-arginine to L-citrulline by the enzyme NO synthase (NOS). There are inducible (i) and constitutive forms of the enzyme. The generation of NO by iNOS (NOS II) occurs after exposure to proinflammatory cytokines, and the levels of NO produced are sustained and in the nanomolar range. In the lung, cell sources of iNOS include airway epithelium, macrophages, monocytes, and eosinophils. The constitutive forms of NOS are found in neurons (nNOS, NOS I) and endothelial cells (eNOS, NOS III). In contrast to iNOS, the release of NO from nNOS and eNOS occurs in bursts, and NO levels are in the picomolar range (1).
Studies evaluating the role of NO on mast cell/basophil activation using NO donor drugs in vitro indicate that NO can inhibit degranulation and cytokine production (24). These findings are consistent with in vivo studies indicating that NO donors can inhibit mast cell degranulation, mast celldependent granulocyte adhesion, and microvascular leakage (5). In addition, NO inhibits activation of nuclear factor- The vascular endothelium releases NO through the action of eNOS in response to blood flowinduced changes in laminar and pulsatile shear stress (1416). Adams and colleagues (17, 18) described a noninvasive motion platform that oscillates an animal in a foot-to-head direction (z-plane) along the spinal axis. The platform varies the frequency and intensity of acceleration in a controlled fashion, imparting periodic sinusoidal inertial forces (pGz) to the body that add pulses to the circulation as the body accelerates and decelerates. Subsequent studies in healthy human subjects (19, 20), in patients (20, 21), in whole animals in vivo (15, 22), and in isolated vessels in vitro indicate that application of pGz results in the generation of eNO (15); consistent with these findings is the observation that the effects of pGz can be blocked by L-NAME (15). Furthermore, in vivo studies demonstrate that the pGz-induced release of eNO provides protection in different models of hemodynamic inflammation (17, 23, 24). Because pGz activates eNOS by the addition of pulses to the circulation throughout the entire body and the diffusion distance for NO from the microcirculation has a range of 150300 µm within 415 s (25), eNO has the potential to act on mast cells and smooth muscle cells within the airways. Although the previous study in patients found no effect on airway responses with pharmacologic blockade of NO during antigen provocation (12), that study did not evaluate whether increased concentrations of NO released by activation of eNOS could modulate allergic airway responses. Based on previous work demonstrating that pGz can increase eNO (15, 1922), this question can now be addressed by using pGz to stimulate eNOS to cause the release of eNO in vivo and determining the effects on antigen-induced airway responses.
The sheep has been used extensively to study the pathophysiology of asthma (26). A primary strength of the model is the ability to repeatedly and accurately assess changes in measures of pulmonary function after provocation with antigen and other irritant stimuli. These functional changes have been characterized extensively and are similar to the respective pathophysiologic changes seen in human subjects in clinical studies (26). Therefore, in this study we tested the hypothesis that pGz could modulate allergen-induced airway responses by examining the effects of pGz treatment on allergen-induced early (EAR) and late (LAR) airway responses, postantigen-induced AHR, and airway inflammatory cell recruitment in allergic sheep. Studies were performed in the presence and absence of L-NAME. Our results show that pGz can block these antigen-induced airway responses and that this protection is reversed by L-NAME. Mechanistically, the protective effects of pGz seem to be mediated, in part, through suppression of NF-
Detailed methods are contained in the online supplement.
Animal Preparation
Whole-body pGz For control studies, the animals in carts were placed on the platform for the appropriate time without motion.
Airway Mechanics
Concentration Response Curves to Carbachol Aerosol
Bronchoalveolar Lavage
NF-
Experimental Protocols To determine if pGz treatment induced a generalized stress response, we measured serum cortisol levels before treatment, 15 min after a 1-h pGz treatment, and 6 h after treatment in six sheep. Studies were done in pairs with control sheep (no pGz) positioned next to the treated animals. The same study was performed in six L-NAME + pGztreated animals and six L-NAMEtreated control animals (no pGz). Sampling times were based on previous work showing that sheep responded to a stress response with an immediate and sustained elevation in systemic cortisol (34).
Protocol 2: Effect of L-NAME on antigen-induced responses.
Protocol 3: Effect of acute pGz treatment when given after antigen challenge on LAR and AHR.
Protocol 4: Effect of multiple pGz treatments on antigen-induced responses.
Protocol 5: Effect of acute pGz treatments with and without L-NAME on BAL cell NF-
Statistical Analysis
Effects of Acute pGz on Antigen-induced Responses (Protocol 1) Figure 1A illustrates the time course of the response to antigen with and without pGz treatment in the presence and absence of L-NAME, and Figure 1B illustrates the effects of these treatments on antigen-induced AHR. pGz significantly reduced the EAR and blocked the LAR when compared with control (Figure 1A), but pretreating the animals with L-NAME abolished this protection (Table 1). We assessed the area under the curve (AUC, % change RL/h) for the EAR (04 h) and LAR (48 h). During the control trial, the EAR AUC was 811 ± 67, and the LAR AUC was 368 ± 26. Treatment with pGz reduced the EAR AUC by 61 ± 5% to 302 ± 32 (p < 0.05) and reduced the LAR AUC by 77 ± 4% to 82 ± 8% (p < 0.05). These effects were reversed by treating the sheep with L-NAME before applying pGz. In the L-NAME + pGz arm, the EAR AUC and LAR AUC were 735 ± 53 and 358 ± 13, respectively. These values did not differ from the control responses but were significantly different (p < 0.05) from pGz alone. The protective effects of pGz on these antigen-induced responses were not related to a change in bronchial tone because there was no difference in post-treatment RL among the three trials (Table 1).
Despite the protection against the antigen-induced bronchoconstrictor responses, the single treatment with pGz did not protect against the post-challenge AHR (Figure 1B, Table 1). In the control trial, PC400 at 24 h fell to 10.8 ± 1.2 BU from a prechallenge value of 22.6 ± 1.8 BU (p < 0.05), resulting in a decreased post-/prechallenge PC400, which is indicative of AHR. When the animals were treated with pGz, the post-/prechallenge PC400 showed a similar decrease, as the prechallenge PC400 value 23.0 ± 2.1 BU fell to 11.6 ± 2.0 BU after challenge. The degree of AHR in the L-NAME + pGztreated animals was similar to the other two arms (Figure 1B, Table 1). Although pGz had no effect on basal airway tone, the protective effects on the antigen-induced EAR and LAR with pGz could have resulted from a decrease in airway smooth muscle responsiveness. To check this, we measured PC400 before and after a 1-h treatment with pGz. PC400 was 30.6 ± 0.4 BU before and 30.6 ± 1.6 BU after a 1-h pGz treatment, indicating that pGz treatment did not affect airway smooth muscle responsiveness (individual responses are presented in Figure E1 in the online supplement). To ensure that the protective effect of pGz on the antigen-induced EAR and LAR was not related to the induction of a stress response in the animals, we compared serum cortisol levels in pGz-treated and control sheep without and with L-NAME pretreatment. Cortisol levels in pGz-treated and control animals in the absence or presence of L-NAME showed a similar profile with cortisol levels declining throughout the day (Figures 2A and 2B). More importantly, there was no evidence of a cortisol surge in the pGz-treated sheep, suggesting that the protective effect of pGz was not due to a stress-related increase in serum cortisol.
Effect of L-NAME on Antigen-induced Responses (Protocol 2) Consistent with the previously reported human data, treatment with L-NAME alone had no effect on the antigen-induced EAR, LAR, or AHR (Figures 3A and 3B, Table 1). These data, in addition to those from the first series of experiments, indicate that the pGz-induced protection is mediated, at least in part, through an NOS-dependent pathway.
The Effect of pGz on LAR Is Not Dependent on Blocking EAR (Protocol 3) To determine if the effects on the LAR seen in Figure 1A were dependent on the reduction in the EAR, animals were treated with pGz between 2 and 4 h after antigen challenge, and the effects on the LAR and AHR were determined. The antigen-induced peak EAR in the control and treatment trials were not different, but pGz treatment provided 81% protection against the LAR (p < 0.05) (Figure 4, Table 1). Nevertheless, as was seen in the pretreatment study, the posttreatment effect of pGz did not affect the antigen-induced AHR (Figure 4, Table 1).
Multiple Treatments with pGz Block Antigen-induced AHR (Protocol 4) To determine if the effects of pGz could be extended so that treatment would also block the antigen-induced AHR, we used a 4-d treatment paradigm, which has been used in previous pharmacologic studies to provide more prolonged and improved efficacy against these antigen-induced airway responses (30). When sheep were treated with pGz twice a day for 3 d and then 1 h before antigen challenge on Day 4, there was protection not only against the antigen-induced EAR and LAR but also against AHR (Figure 5, Table 1). In the control trial, the EAR AUC and the LAR AUC were 872 ± 36 and 374 ± 33, respectively. When these animals were treated with pGz, the EAR AUC and the LAR AUC were significantly reduced (p < 0.05) to 283 ± 8 and 52 ± 11, respectively. The degree of inhibition of the EAR (68 ± 1%) and LAR (87 ± 2%) seen with the extended pretreatment time was not different from that seen with the 1-h pretreatment. Nevertheless, the effect on AHR was evident because the PC400 after challenge (23.3 ± 2.3 BU) was not different from the prechallenge value of 24.0 ± 2.8 BU. This contrasts with the control response where PC400 after challenge fell to 12.0 ± 0.6 BU from a prechallenge value of 24.5 ± 2.1 BU (p < 0.05).
To determine if the multiple treatments with pGz affected the antigen-induced leukocyte infiltration into the airways, we compared the changes in the BAL cell response after antigen challenge in multiple-pGztreated and control animals. Animals that received multiple pGz treatments had a significantly lower BAL neutrophilia (5.7-fold) compared with untreated animals (Figure 6). Total cells (2.7-fold) and lymphocytes (2.4-fold) were also reduced in the treated animals, but the relative changes were not significantly different between the groups.
pGz Reduces Antigen-induced Increases in NF- B Activity in BAL Cells (Protocol 5)The ability of acute pGz treatments to block the LAR and multiple treatments to block the LAR and the AHR and to reduce the inflammatory cell influx after antigen challenge suggests that pGz modulates inflammatory pathways. NF- B is a key transcription factor regulating inflammatory gene expression of proinflammatory cytokines and adhesion molecules. Because eNO is reported to inhibit NF- B (8) and because pGz stimulates eNOS, we reasoned that NF- B activity in animals treated with pGz should be lower than in untreated animals after antigen challenge. Furthermore, if pGz reduced NF- B activity through an NO-dependent pathway, this effect should be abolished by L-NAME. To test this hypothesis, we repeated the antigen challenge studies with and without a 1-h pGz treatment in the presence or absence of L-NAME, except that the sheep were lavaged immediately after the 6-h measurement of RL. The recovered cell pellet was assayed for NF- B activity. For this study, the respective peak EAR and LAR (at 6 h) for the control group was 481 ± 38% and 118 ± 3%; as shown in the previous study, the EAR (222 ± 27%) and LAR (21 ± 4%) were significantly reduced in the pGz-treated animals, and this protection was lost in the L-NAME + pGztreated animals (EAR 598 ± 64% and LAR 130 ± 4%). Figure 7 shows that the cells recovered from the control group and the pGz + L-NAME group had 1.9- and 1.8-fold increases (p < 0.05), respectively, in NF- B activity over the pGz-treated sheep. Thus, pGz not only blocked the functional airway effects and the inflammatory cell response to antigen challenge but also was a key inflammatory signaling factor. This protection was lost in the presence of L-NAME.
This study provides the first evidence that whole-body pGz can modulate antigen-induced functional airway responses and airway inflammatory cell recruitment in an animal model of asthma. This protection is blocked by L-NAME, suggesting that the effect of pGz is NO dependent. Because previous studies in animals and humans demonstrate that pGz releases eNO (15, 20, 22) and that, at least in animal studies, this eNO can reduce pathophysiologic responses to inflammatory stimuli (17, 24), one can speculate that the pGz-induced protection observed in the present study occurs through a similar mechanism. Although there is evidence from previous animal and human studies (15, 20) that pGz stimulates eNOS resulting in NO release, our conclusions must remain speculative because direct measurements of eNO were not made in the present study. What is not speculative, however, are the novel findings that pGz inhibits antigen-induced functional and inflammatory responses in the sheep model of asthma and that the extent of the protection is dependent on the number of treatments. The evidence that pGz can stimulate eNOS resulting in NO release is based on multiple sources. It is well known that in response to blood flowinduced changes in laminar and pulsatile shear stress the vascular endothelium releases NO through the action of eNOS (1416). pGz has been shown to increase vascular shear stress, resulting in decreased pulmonary and systemic vascular resistances, which is consistent with the release of NO from endothelial sources. Direct evidence that pGz can stimulate the release of eNO comes from studies in anesthetized, supine piglets, where the pulsatile shear stress produced by 60 min pGz significantly increased serum nitrite, a marker of eNOS activity (35), and attenuated the vascular constrictor responses caused by L-NAME infusion (15). Adams and colleagues (15) confirmed these in vivo observations by applying pGz to an isolated vessel preparation in vitro. Effluents from porcine aortas exposed to pulsatile flow and pulsatile flow plus pGz showed 300 and 1,000% greater increases in nitrite levels, respectively, when compared with nitrite levels in effluents collected from vessels exposed to nonpulsatile flow (15). In human subjects, Sackner and colleagues provided indirect evidence that pGz released NO by measuring the extent of the descent of the dicrotic notch down the diastolic limb of the finger pulse (19, 20), a measure that reflects the vasodilator action of NO on the resistance vessels owing to the delay of pulse wave reflection (36, 37). This study further showed that the eNOS-stimulating effects of pGz were maintained throughout the 45 min of treatment (20), which, if extended to the current study, would translate to pGz stimulating eNOS activity throughout the 1-h treatment time. The speculation in the present study that pGz stimulates release of NO from eNOS is also based on the pharmacologic abolition of the protective effects of pGz with L-NAME. This, in conjunction with the finding that L-NAME does not affect antigen-induced responses in non-pGztreated animals, a result consistent with previous findings in patients with asthma (12), suggests that the NO generated by pGz is critical to the protective effects seen in our study. It is unlikely that iNOS was involved in the pGz protective effect because iNOS is thought to contribute to, but not block, antigen-induced inflammatory processes (1, 38). Furthermore, in patients with asthma, L-NAME significantly reduced exhaled NO, which is thought to reflect iNOS activity, but not the EAR or LAR. These observations led to the conclusion that iNOS expression and elevated iNO are not causally related to antigen-induced bronchoconstriction and that increased iNOS activity and iNO levels are a consequence of, rather than the cause of, the LAR and subsequent inflammatory response. In the present study, we extend the observations made in subjects with asthma and non-pGztreated allergic sheep that L-NAME has no effect on antigen-induced responses by showing that, in pGz-treated animals, L-NAME reverses the protective effect of pGz. These findings indicate that the pGz-stimulated NO, which we suggest is due to increased eNOS activity, is responsible for the observed beneficial effects and that this NO is different from that in untreated animals. In the presence of L-NAME, this pGz-dependent NO is blocked, resulting in antigen-induced airway responses that are superimposed with the control (untreated) responses. L-NAME is widely used to block NO synthesis in vivo and in vitro (39), and although it is not a specific inhibitor of eNOS, L-NAME has a preference for eNOS over iNOS (40). Our findings also make it unlikely that nNOS was stimulated by pGz. Neuronal NOS is thought to be a bronchodilator and to reduce airway responsiveness (1). As seen in our study, pGz had no effect on baseline tone or baseline bronchial responsiveness, which suggests that nNOS was not affected by pGz. Collectively, this evidence supports our contention that the pGz-induced effects are mediated through eNOS, a conclusion that is consistent with the reported beneficial antiinflammatory actions of eNO. The primary aim of this study was to determine if pGz could modify allergen-induced airway responses in allergic sheep. This model is well suited for these studies because one is able to repeatedly and accurately assess changes in measures of pulmonary function after provocation with antigen (26). The functional changes (i.e., the EAR, LAR, and AHR) are similar to the respective pathophysiologic changes seen in human subjects in clinical antigen challenge studies, and the mechanisms responsible for the antigen-induced EAR, LAR, and AHR are well described (26). Within this context, two factors are critical for interpreting the present results. First, the mediators recovered from sheep airways in association with the antigen-induced changes in airway function are consistent with those recovered in patients with asthma after allergen challenge (41, 42). Second, when these responses (EAR, LAR, and/or AHR) are blocked pharmacologically, the abrogation of the antigen-induced airway response is accurately reflected by the reduction in relevant mediator levels (4143). NO can inhibit mast cell secretion in vitro and in vivo (3). The NO donor sodium nitroprusside inhibits anti-IgEinduced histamine and tryptase release from human skin mast cells (2) and histamine release from human basophils in vitro (44). In rats, an NO donor inhibited mast cell degranulation, mast celldependent granulocyte adhesion, and microvascular leakage in vivo (5). These latter findings are consistent with studies in pigs in which inhibition of NOS enhanced allergen-induced histamine release (45). Because the EAR in sheep, as in humans, is mast-cell dependent (42, 46, 47) and because pGz has no effect on airway tone and/or airway responsiveness, our findings support the concept that the inhibitory action of pGz on the EAR is through modulation of mast cell activation rather than bronchodilation. The protective effect of pGz was lost if the animals were treated with L-NAME before administering pGz. If we accept the argument that pGz can stimulate eNOS, the protective effects of pGz are consistent with the reported ability of NO to inhibit mast cell/basophil histamine and tryptase release (2), protective actions that would be lost in the presence of L-NAME. The development of antigen-induced LAR is the initial sign of a heightened inflammatory process that leads to a prolonged increase in AHR (48). Acute pretreatment with pGz blocked the antigen-induced LAR, an effect that could be a linked to the initial reduction in the EAR. However, the finding that pGz affects the LAR independent of the EAR (Figure 4) suggests that pGz has antiinflammatory effects as well. Our previous work demonstrates that the LAR can be blocked by adhesion molecule inhibitors and/or specific mediator antagonists when given after the EAR, indicating that interfering with the recruitment of activated inflammatory cells to the airways is important for the development of the LAR (29, 30, 32, 49). That pGz can block the LAR when given after the EAR is consistent with reports that NO can block mast celldependent inflammatory processes (3), IgE-dependent cytokine production (4), and inflammatory cell recruitment and adhesion (5). Although acute pretreatment and/or acute post-treatment with pGz protected against the antigen-induced EAR and/or LAR, both acute treatments were ineffective in preventing the postantigen-induced AHR. The lack of protection for the latter can be attributed to the fact that a single pGz treatment is unlikely to sustain a sufficient signal to counteract the AHR that occurs 24 h after challenge. Therefore, in an attempt to extend the protective effects of pGz, we used a treatment paradigm that we demonstrated to be successful in blocking antigen-induced airway responses with other pharmacologic agents (30, 50). By using multiple pretreatments on the days before antigen challenge, we uncovered increased efficacy of compounds that were marginally effective and/or ineffective with acute dosing (30, 50). This treatment paradigm was successful in the present study because multiple twice-daily treatments with pGz successfully blocked the AHR that occurred 24 h after antigen challenge. In the Ascaris-induced sheep model of asthma, the neutrophil is the most prominent leukocyte found in BAL fluid after antigen challenge, which signals an active inflammatory process (3032, 51, 52). In the present study, multiple pGz treatments reduced this neutrophil influx into the airways; this finding is consistent with adhesion molecule inhibitors that have blocked the inflammatory cell influx and antigen-induced AHR (3032, 51, 52). In spite of the extended protective effect seen with multiple pGz treatments over 4 d, there was no statistically significant improvement in the degree of inhibition of the EAR or LAR. The reasons for this are not clear, although one could speculate that because the level of pGz was the same, the resultant stimulation of eNOS was similar in the two treatment regimens, but with the chronicity of multiple treatments, different molecular events, including the up-regulation of eNOS protein levels (14) and/or the increase the cellular S-nitrosylation, could occur (53). Both of these mechanisms could theoretically contribute to a prolonged release of NO (14).
Increased NF- Dogs treated with the cortisol synthesis inhibitor metyrapone were found to develop LAR after airway challenge with A. suum, suggesting that elevations in systemic cortisol levels could modulate the LAR response (63). In sheep, Apple and colleagues (34) found an immediate and sustained rise (up to 6 h) in plasma cortisol after the initiation of stress (leg restraint). Thus, if pGz were initiating a generalized stress response, it should be apparent immediately after treatment and should be sustained over the time course of our studies. Based on the dog study (63), such a rise in systemic cortisol could explain the protective effects of pGz on the LAR. No such stress response was evident in the pGz-treated or nontreated sheep in the presence or absence of L-NAME (see Figure 2). Therefore, it is unlikely that the protective effects of pGz can be attributed to a generalized stress response in these animals. In conclusion, our findings indicate that pGz has beneficial effects on the pathophysiologic airway responses and inflammatory cascades associated with allergen challenge in an experimental model of asthma. These observations, if confirmed in patients with asthma, suggest that chronic pGz treatment, a noninvasive modality, could provide an alternative medicineadjunctive therapy approach that could be beneficial as a means of asthma control.
The equipment used in this study was provided by Acceleration Therapeutics.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200601-0480OC on July 20, 2006 Conflict of Interest Statement: W.M.A. does not have a financial relationship with a commercial entity that has an interest in the subject of the manuscript. A.A. does not have a financial relationship with a commercial entity that has an interest in the subject of the manuscript. I.S. does not have a financial relationship with a commercial entity that has an interest in the subject of the manuscript. I.T.L. does not have a financial relationship with a commercial entity that has an interest in the subject of the manuscript. J.B. owns shares of Non-Invasive Monitoring Systems, Inc. (NIM) stock. J.A.A. owns shares of NIM stock, which manufactures AT-101. M.A.S. is the chairman of the board of directors of NIM. He received no salary in 2004 and $26,000 in 2005. He holds approximately 22% shares of the company. Received in original form January 12, 2006; accepted in final form July 3, 2006
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||