INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ANIMAL MODELS OF ASTHMA HAVE been developed in various animal species and have provided new insights into the mechanisms underlying the pathophysiology of airway hyperresponsiveness in asthma. The highly inbred Brown-Norway (BN) rat model has been widely used to study respiratory mechanics in allergen-induced airway inflammation because of its well-developed airway smooth musculature, within-strain homogeneity of airway reactivity, and the development of early and late airway responsiveness (AR) to allergen (6, 7, 27). Recent advances in measuring respiratory mechanics in intact, spontaneously breathing animals have allowed simple, noninvasive, routine analysis of pulmonary function in animal models (11, 12, 22, 23, 26). Results of our previous studies suggest that midexpiratory tidal flow (EF₅₀), as measured by head-out body plethysmography, can be used as a noninvasive measure of bronchoconstriction in mice (9, 21).¹ The assessment of EF₅₀ in conscious animals has a number of advantages. First, the measurements are technically easy to perform. Second, because this method does not require anesthesia or surgical interventions, the same animal can be studied on multiple occasions. Until now, only few data on the validity of EF₅₀ in mice have been available. However, it is unclear whether this parameter can simply be transferred to other species with a different breathing pattern and whether this index is suitable to detect changes associated not only with ACh-related bronchoconstriction but also with allergen-specific early and late AR. Accordingly, the reliability of EF₅₀ has to be studied more extensively, in direct comparison with conventional standard parameters, before it can be established as a surrogate for bronchoconstriction in experimental asthma models.

The aim of this study was to provide first experimental evidence that the noninvasive determination of EF₅₀ can be used as a valid physiological measure of nonspecific and allergen-specific AR in conscious rats. To investigate the applicability of EF₅₀ in a rat model of early and late AR, dose-response studies to aerosolized ACh were performed with noninvasive measurement of EF₅₀ in conscious rats by head-out body plethysmography, followed by invasive monitoring of simultaneously measured EF₅₀ and pulmonary conductance (GL) in orotracheally intubated, spontaneously breathing rats. The sensitivity of EF₅₀ to detect early and late AR in allergic rats was compared with the "gold standard" parameter, GL. In addition, we examined the potential effects of induced hyper- and hypoventilation on simultaneously measured EF₅₀ and GL. This approach represents a significant advance in accuracy over previous validation experiments that have failed to demonstrate simultaneous measurements of noninvasive and invasive indexes of bronchoconstriction or that have included upper airway resistance (9, 11, 12, 21).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Pathogen-free inbred male BN rats (Charles River, Sulzfeld, Germany), 6-7 wk of age, were used in all experiments. Rats were housed for at least 1 wk before investigation. Rats were divided into four main groups: 1) ACh challenge (n = 8) (after noninvasive monitoring of EF₅₀, animals were reused for invasive determination of EF₅₀ and GL to ACh challenge 24 h later); 2) challenge experiments with saline, CO₂, and halothane (n = 5); 3) allergen-sensitized and boosted rats for invasive (n = 8) and noninvasive (n = 8) measurements of early and late AR to inhaled ovalbumin (Ova); and 4) sham-treated controls for invasive (n = 6) and noninvasive (n = 6) measurements of early and late AR to inhaled Ova. All animal experimentation conformed to the guidelines of the National Research Council Guide.

Sensitization protocol. For allergic rats (Ova/Ova), on day 0, animals were sensitized by a subcutaneous injection of 1 mg Ova (grade V, Sigma Chemical, Deisenhofen, Germany) precipitated in 20 mg of aluminum hydroxide gel in 0.25 ml normal saline (n = 8). An adjuvant consisting of 6 × 10⁹ heat-killed Bordetella pertussis organisms (Chiron-Behring, Marburg, Germany) was injected intraperitoneally at the same time. On days 7 and 14, the anesthetized animals were boosted intratracheally via an orotracheal tube with 0.1 ml of 0.3% Ova in normal saline (NaCl). Control rats (NaCl/NaCl) were sensitized and boosted to normal saline instead of Ova (n = 6). On day 21, all animals were challenged with aerosolized Ova as described below.

Noninvasive measurement of AR in conscious rats. AR was assessed as previously described in mice (9, 26) with a modified head-out body plethysmography system for rats (head-out body plethysmography system model 855, HSE-Harvard, March-Hugstetten, Germany). Briefly, rats were placed in the body plethysmographs while the head of each animal protruded through a neck collar (18-20 mm ID, dental latex dam; Roeko, Langenau, Germany) into the head-exposure chamber, which was ventilated by a continuous airflow of 0.5 l/min. Rats held in head-out body plethysmographs tolerated noninvasive measurements very well. Monitoring of respiratory function was started after 2-5 min when animals and individual measurements settled down to a stable level. For airflow measurement, a wire mesh pneumotachometer with defined resistance (6 layers of wire mesh cloth, linearity ±2% ranging from 20-300 breaths/min; HSE-Harvard) and a differential pressure transducer (Validyne DP 45-14, range ±2 cmH₂O, HSE-Harvard) coupled to an amplifier were attached in midposition to the side wall of each plethysmograph. For each animal, the amplified analog signal from the pressure transducer was digitized via an analog-to-digital converter (DT 302, Data Translation) at a sampling rate of 250 Hz. Before each measurement, the plethysmograph was calibrated with an injection of 1 ml air into the body chamber. The pneumotachograph tidal flow signal was integrated with time to obtain tidal volume (VT). From these signals, all of the noninvasive parameters listed in Table 1 [EF₅₀, time of inspiration (TI) and expiration (TE), VT, respiratory rate (f)] were calculated for each breath with commercial software (HEM 3.3, Notocord, Paris, France).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Left: normal breathing pattern of an anesthetized, orotracheally intubated, spontaneously breathing Brown-Norway rat. Right: breathing pattern during bronchoconstriction due to inhalation of ACh aerosol, illustrating the simultaneous decreases in midexpiratory tidal flow (EF50) and pulmonary conductance (GL). A: tidal volume (VT) obtained from the integrated airflow signal over time. A vertical line indicates the value of EF50 at midexpiratory tidal volume. B: corresponding airflow signal from the pneumotachograph during expiration (above zero) and inspiration (below zero). C: corresponding GL signal.

Invasive determination of AR. AR was assessed in anesthetized and orotracheally intubated rats as previously described (14-16). The spontaneously breathing animal was placed in the supine position in a thermostat-controlled whole body plethysmography, and slight anesthesia was maintained with a gas mixture of 0.8% halothane in 30% oxygen. Tidal flow was determined by a pneumotachometer with defined resistance (8 layers of wire mesh cloth; HSE-Harvard) attached to the rear wall of the plethysmograph and connected to a differential pressure transducer (Validyne MP 45-1, range ±2 cmH₂O; HSE-Harvard). For measurement of esophageal pressure, a water-filled polyvinyl chloride tube (13.5-cm length, 1-mm ID) was inserted into the lower one-third of the esophagus and coupled to a pressure transducer (model P75, range ±100 cmH₂O; HSE-Harvard). The amplified analog signal from the pressure transducer was digitized by an analog-to-digital converter (DT 302, Data Translation) at a sampling rate of 250 Hz. Pulmonary resistance (RL) was calculated from the integrated differences in esophageal pressure and tidal flow (14-16). RL and EF₅₀, together with all of the other parameters listed in Table 1, were recorded simultaneously with commercial software (HEM 3.3, Notocord). For better comparability of data, RL was expressed as GL (1/RL).

Challenge protocol. After recording of baseline values, AR to doubling doses of aerosolized ACh (ranging from 20 to 160 mg/m³; Sigma Chemical) in naive rats and AR to aerosolized Ova (grade V, 135 mg/m³ over 8.5 min) in sensitized and nonsensitized animals were assessed. Aerosols were generated by a computer-controlled aerosol generator system (Bronchy III, particle size 1.5-2 µm mass median aerodynamic diameter; Fraunhofer ITA, Hannover, Germany), as previously described (14-16). To characterize the potential effects of induced hyper- and hypoventilation on the parameters of bronchoconstriction, inhalational exposures to 10% carbon dioxide (CO₂) and 3.5% halothane (Eurim-Pharma, Piding, Germany) were performed. The degree of AR to all of these challenges was determined from minimum values of GL and EF₅₀ and was expressed as percent changes from corresponding baseline values, which were taken as 0%.

Determination of total IgE in serum. Plates were coated with mouse monoclonal antibody to rat IgE (MCA 193, Serotec, Wiesbaden, Germany) in a concentration of 0.1 µg/ml per well. The plates were incubated over night, washed, blocked with Roti Bock (Roth, Karlsruhe, Germany), and incubated with sera and standard for 1 h. Total IgE protein (rat IgE, PRP07, Serotec) was used as a standard in a concentration of 0.2 µg/ml. After washing, 50 µl of a second peroxidase-labeled anti-IgE antibody (MCA 187P, Serotec) were added to the plates and incubated for 1 h. After an additional washing step, 50 µl of the peroxidase substrate o-phenylenediamine dihydrochloride (Sigma Chemical) were pipetted to each well. The plates were developed for 10 min, and the reaction was stopped by adding 100 µl/well of 1 N HCl (Merck, Darmstadt, Germany). Plates were read with an ELISA plate reader (model MRX, Dynatech, Chantilly, VA) at 490 nm.

Assessment of leukocyte distribution in bronchoalveolar lavage fluid and histology. Bronchoalveolar lavage (BAL) and histology were performed 16-20 h after allergen challenge, as previously described (25).

Statistics. A mixed model was fitted to the data by using the animal as random factor (4). Comparisons for multiple pairs were performed by the Tukey Kramer test and for single pairs by Student's t-test. Correlations were calculated with Pearson's correlation coefficient. Values for all measurements were expressed as means ± SD. The agreement between the methods was analyzed by the method of Bland and Altman (3). Agreement was expressed as the mean differences obtained by the different techniques (e.g., between the EF₅₀ method in conscious and anesthetized rats and between EF₅₀ and GL in anesthetized animals). The limits of agreement were expressed as means ± 2 SD, and the 95% confidence intervals (CIs) of the mean as well as the lower and upper limits of agreement were calculated according to Bland and Altman. P values < 0.05 were considered significant. All statistical analyses were performed with SAS 8.01.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Baseline values for respiratory parameters in conscious and anesthetized BN rats. Table 1 presents the baseline values of respiratory parameters obtained from naive, conscious BN rats placed in head-out plethysmographs and from anesthetized, spontaneously breathing BN rats placed in whole body plethysmographs. Except for EF₅₀, anesthesia significantly affected the respiratory parameters f, VT, TI, and TE compared with baseline levels of conscious animals.

Comparison of the noninvasive determination of EF₅₀ with invasive measurements of GL and EF₅₀ in AR to ACh. Nonspecific AR to increasing doses of aerosolized ACh (20-160 mg/m³) was investigated in naive, conscious rats (n = 8) with noninvasive head-out body plethysmography. The same animals were anesthetized and invasively reexposed to inhaled ACh 24 h later. Figure 1 presents the breathing pattern of an anesthetized, spontaneously breathing rat that was breathing mixed air (room air with 30% oxygen) and the characteristic modifications to the normal airflow pattern after ACh aerosol exposure, illustrating the simultaneous decreases in VT, EF₅₀, and GL during bronchoconstriction.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Dose-response relationship to aerosolized ACh (20-160 mg/m3) in naive animals. Noninvasive determination of the decline in EF50 to ACh was followed by invasive recording of simultaneously measured decreases in EF50 and GL to ACh exposure in the same animals 24 h later. EF50 and GL were allowed to return to baseline before each subsequent challenge. Values are means ± SD (n = 8 rats) of percent changes to corresponding baseline values, which were taken as 0%. Baseline values for EF50 and GL were within the means ± SD as listed in Table 1. No significant differences in dose-related changes were observed between noninvasively and invasively measured EF50.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Individual differences in the amount of ACh-induced bronchoconstriction between simultaneously measured EF50 and GL in anesthetized (an) rats (A) and between EF50 in conscious (con) and anesthetized animals (B) are plotted against the average corresponding values (expressed as %change to baseline). Solid line, mean of the differences; dashed lines, upper and lower limits of agreement (±2 SD).

Effects of CO₂ and deep anesthesia on respiratory parameters. To investigate the influence of various agents with potential modifying effects on f, VT, EF₅₀, and GL recordings, a separate group of naive, anesthetized rats (n = 5) was exposed to 10% CO₂ and deep anesthesia with 3.5% halothane (Table 2). Except for a stimulating effect on f and VT, CO₂ exposure evoked no significant decreases in EF₅₀ and GL. Standard anesthesia with 2% halothane induced a reduction in f and VT (see Table 1). Changing halothane concentrations from 2 to 3.5% resulted in an additional decline in f and VT, as well as in significant decreases in EF₅₀ and GL values. These experiments demonstrate that EF₅₀ was not decreased by CO₂ exposure and that the changes in EF₅₀ to 3.5% halothane exposure were quite similar to those simultaneously observed with GL.

Early and late AR to Ova in allergic rats. The early and late AR to Ova was investigated on day 21 in two groups divided into allergen- (Ova/Ova) and saline-treated (NaCl/NaCl) animals (Fig. 4). To avoid unbalanced challenges with Ova, each group was separated into two subgroups for invasive and noninvasive measurement of pulmonary function. All respiratory parameters as listed in Table 1 were continuously monitored during the aerosol challenge with Ova.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Assessment of early airway responsiveness (AR; A) and late AR (B) to ovalbumin (Ova) challenge (concentration = 135 mg/m3). AR to Ova was monitored in Ova/Ova rats (n = 8 each) and in NaCl/NaCl animals (n = 6 each). The EF50 responses were studied in separate groups of invasively and noninvasively monitored animals. Both early AR and late AR were related to prechallenge baseline values. Significant decreases in invasively and noninvasively recorded EF50 and GL within 23 min and at 5-6.5 h after Ova challenge were observed in Ova/Ova rats. AR is expressed as %change in EF50 and GL from corresponding baseline values, which were taken as 0%. Baseline values were within the means ± SD as listed in Table 1 and were similar in the Ova and saline group. * P < 0.01 (early AR) and ** P < 0.05 (late AR) for each parameter vs. NaCl/NaCl.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Characteristic time course of an early AR illustrating the parallel decreases in EF50 and GL values with a concomitant slight decline in VT during and after aerosolized Ova challenge of an allergic rat. There was an evident recovery of EF50 and VT during the postchallenge period, which was less pronounced for GL. AR is expressed as %change in VT, EF50, and GL from baseline values, which were taken as 0%. Baseline values were within the means ± SD as listed in Table 1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In this report, we describe that the changes in EF₅₀ closely reflected the changes observed in GL during bronchoconstriction induced by aerosolized ACh and allergen in rats. We have characterized airway responses to ACh in naive animals and to allergen in a rat model that replicates the dual-phase response in asthma.

The increasing amount of genetic information from experimental asthma models has extended our understanding of the complex pathophysiology of asthma. However, there is a large discrepancy between the availability of the advanced tools of cell and molecular biology and the conventional invasive methods of assessing lung function in animal models (1). Accordingly, the development of an accurate and sensitive methodology that allows simple and repetitive determination of respiratory physiology in the intact organism is necessary to study the functional consequences of in vitro findings in the lung in vivo. Our strategy of noninvasive determination of EF₅₀ with head-out body plethysmography enables both repetitive assessment and quantitative interpretation of bronchoconstriction in conscious, spontaneously breathing animals.

The values for respiratory parameters monitored in conscious and in anesthetized rats were reproducible and paralleled those reported previously for male BN rats (Table 1) (28). Standard anesthesia with halothane did not influence baseline EF₅₀ measurements compared with those in conscious animals (see Table 1). To evaluate the reliability of EF₅₀ as a measure of bronchoconstriction, we performed dose-response measurements to inhaled ACh, ranging from 20 to 160 mg/m³, in both conscious and anesthetized animals. In these comparative studies, we demonstrated that EF₅₀ accurately reflected the response to increasing doses of ACh observed with GL as an outcome indicator. Direct comparison of invasively recorded EF₅₀ and GL with noninvasive measurement of EF₅₀ revealed that GL agreed closely with invasive and noninvasive EF₅₀ recordings and showed comparable decreases of the dose-response curve over baseline values (Figs. 2 and 3). Consistent with data from previous studies (2, 6, 7, 27), we found that all BN rats immunized and boosted by Ova uniformly demonstrated an early AR followed by a late AR with a prevalence of 75%. This dual-phase AR was allergen specific, as the extent of both the early and late airway responses after provocation with Ova was significantly lower in control animals (NaCl/NaCl). It should be pointed out, however, that the post-Ova decrease in EF₅₀ might relate to diminished lung compliance caused by heterogeneous airway closure and non-reopening after Ova-induced bronchoconstriction. Detection of early- and late-phase AR was based on corresponding decreases in GL and EF₅₀. We found that, in response to ACh or allergen, invasively and noninvasively recorded EF₅₀ tended to slightly underestimate the degree of bronchoconstriction and was sustained for a shorter period compared with the GL method. A possible explanation for this observation might be that GL rather than EF₅₀ reflects changes in tissue resistance, known to largely contribute to the overall resistance to nebulized bronchoconstrictors in rats (10, 20). However, we have demonstrated that EF₅₀ can significantly discriminate the degree of AR and showed acceptable overall agreement with GL recordings. The range of variations from single measurements was not larger than those reported from other animal and human studies comparing invasive and noninvasive techniques (11, 13). In combination with the noninvasive monitoring of EF₅₀ in conscious rats, this approach seems to be more appropriate than previous validation experiments of noninvasive technologies. Compared with the established noninvasive whole body plethysmography method (5, 11, 19) that uses the dimensionless variable enhanced pause (Penh) to empirically monitor bronchoconstriction in conscious animals, the absolute value of EF₅₀ has physical meaning and enables a direct quantitative comparison from animal to animal (5). Consistent with reported data in mice (9, 18), we provide evidence that, in rats also, bronchoconstriction is clearly associated with a reduction in VT (Fig. 5). In contrast, VT has been shown to increase during bronchoconstriction with the Penh method (11). Although methodological differences exist between the Penh and the EF₅₀ methods, including the limited quantitative and physiological interpretation of Penh, it has been shown recently in mice that both methods provide similar and reproducible results (8).

The independence of EF₅₀ measurements from breathing frequency was examined in experiments with 10% CO₂ and 2% halothane, which demonstrated that changes in breathing frequency induced no significant decrease in EF₅₀ and GL values (Tables 1 and 2). However, deep anesthesia with 3.5% halothane was associated with concomitant decreases in f, VT, EF₅₀, and GL values. This finding contrasts with the fact that volatile anesthetics such as halothane usually have direct relaxant effects on airway smooth muscle. The observed bronchoconstrictive response to high concentrations of halothane may be related to a pronounced increase in the tissue component of resistance caused by viscoelastic elements and lung inhomogeneity in the rat (24).

In our model of allergic airway inflammation, decreases in EF₅₀ were linked with enhanced IgE serum levels and the development of neutrophil and eosinophil influx in the lungs after allergen challenge of allergic rats. The ability to measure pulmonary physiology repeatedly in a longitudinal study design would be valuable for the study of the time course of chronic lung and airway disorders. The recent development of simple testing methods in conscious, intact animals has contributed largely to a widespread and multidisciplinary use of pulmonary function tests in experimental asthma research (8, 17, 19).

In summary, EF₅₀ measurements closely reflected the changes in invasively monitored EF₅₀ and GL during challenges with inhaled ACh and allergen. We conclude that the broad applicability of noninvasive measurement of EF₅₀ by head-out body plethysmography should provide a complementary methodology to perform simple and physiologically valid routine analysis of pulmonary function in rat and mouse models of asthma.