INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Eosinophils are recognized as the most important effector cells producing allergic inflammatory airway diseases such as asthma (1). Histologic examination of bronchial biopsy specimens obtained from patients who have asthma shows that a large number of eosinophils and T-lymphocytes are present in the airway wall (2). Increased numbers of eosinophils are also described in sputum and bronchial lavage (5, 6). A wide variety of chemotactic factors and adhesion molecules are postulated to explain the selective accumulation of eosinophils at sites of allergic inflammation (7). However, the mechanism of recruitment, activation, adhesion, and migration of eosinophils to the site of allergic inflammation is still incompletely understood. Because a significant increase of eosinophils in the airway wall is also observed in patients who have nonallergic asthma, chronic obstructive pulmonary disease, or bronchiectasis without an allergic background (8), mechanisms other than an allergic inflammatory process may contribute in the accumulation of eosinophils.

Using morphometric measurements, James and coworkers (9) showed that inner wall area (mucosal area internal to smooth muscle plus smooth muscle area) stays constant during acute bronchoconstriction. Because the mucosal membrane internal to smooth muscle is thrown into folds during bronchoconstriction, the interstitial pressure in the mucosal membrane probably increases (10, 11). Therefore, it is probable that vascular networks in the mucosal membrane deform during bronchoconstriction. We hypothesized that eosinophils are retained in the bronchial vasculature in the inner airway wall during bronchoconstriction because of deformation of the mucosal membrane. To test this hypothesis we measured the eosinophil density in the airway wall of guinea pigs, which received unilateral vagal stimulation by electrical current.

	METHOD

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal Preparation

Male Cam Hartley guinea pigs weighing 250 to 350 g were anesthetized using an intraperitoneal injection of -chloralose (100 mg/kg) and urethane (1,000 mg/kg). Animals were tracheostomized and artificially ventilated with pure oxygen with a tidal volume of 10 ml/kg, respiratory rate of 60 breaths/min, and positive end-expiratory pressure (PEEP) of 4 cm H₂O. The chest was widely opened by splitting the sternum. The left jugular vein was cannulated for the injection of drugs. Tracheal pressure (Ptr) was measured through a side port of the tracheal tube using a piezoresistive transducer (FMP-02PG; Fujikura, Tokyo, Japan) and recorded using a data acquisition system (Raytech, Vancouver, BC). Both cervical vagal nerves were carefully dissected free and sectioned at the neck. Then the right vagus was connected with electrodes of a bipolar electrical stimulator.

Experimental Protocols

Protocol 1. Eight guinea pigs were used. After stabilizing the animal and measuring baseline Ptr, the electrical stimulation (pulse width, 1 ms; frequency, 16 H_z; Voltage; 24 V) was applied only to the right vagus. We selected only the right vagus for the electrical stimulation since we observed a greater bronchoconstriction (increase in Ptr) during right vagal stimulation than during left vagal stimulation in preliminary experiments. Two minutes from the start of electrical stimulation, Ptr was measured, and liquid nitrogen was showered onto the lung and then the whole body was submerged in liquid nitrogen for 20 min to completely freeze the lungs. During these procedures the lungs were kept inflated at a transpulmonary pressure of 4 cm H₂O. Four to eight blocks of lung tissue were dissected from each lung and a freeze substitution technique was used to process the lung for light microscopic examination and morphometric study as described below.

In the preliminary study (n = 4), a catheter was inserted into the carotid artery, and changes in blood pressure (BP) and heart rate (HR) before ad during vagal stimulation was recorded.

Protocol 2. Eight guinea pigs were used. After stabilizing the animals and measuring baseline Ptr, the first electrical stimulation was applied to the right vagus to confirm that bronchoconstriction occurred. The animals that responded to the stimulation (increase in Ptr by 5 cm H₂O or larger) were selected for this protocol. Thirty minutes were allowed for the animals to return to baseline physiologic conditions, and atropine sulphate (1 mg/kg) was injected through a jugular vein. Five minutes after the injection, the second electrical stimulation was applied to the right vagus again for 2 min. The same procedures as in Protocol 1 were used to freeze and process the lung for light microscopic examination and morphometric study.

In a preliminary study (n = 3), we tested the effectiveness of two consecutive vagal stimulations with 30-min interval by comparing increase in Ptr.

Freeze Substitution Technique

The lungs were processed for light microscopic examination using a freeze substitution method as previously described (12). Briefly, the frozen lung pieces were stored in precooled acetone, 70° C for 18 h, 20° C for 6 h, and 4° C for 24 h. Each block was then fixed using 10% buffered formalin, embedded in paraffin, and sectioned at 5 microns. Hansel staining, which stains the basic arginine residues with acidic eosin, was used for identifying eosinophils (13).

Morphometric Measurements

All membranous airways that were cut in reasonable cross section (a short versus a long diameter ratio at the smooth muscle border equal to or greater than 0.6) were examined. Measurements were performed using a Nikon microscope (Nikon Scientific Instruments, Tokyo, Japan) equipped with a camera lucida attachment and a digitizing tablet coupled to an IBM compatible computer. The measurements are illustrated in Figure 1, and they include: (1) the long and short diameters of the outer border of the smooth muscle perimeter (DL and DS); (2) airway internal perimeter (P_i) and area (A_i) (P_i is the perimeter of the luminal border and A_i is the area enclosed by P_i); (3) basement membrane perimeter (P_bm); (4) outer smooth muscle perimeter (P_mo) and area (A_mo) (P_mo is the perimeter of the outer border of the smooth muscle and A_mo is the area enclosed by P_mo); (5) airway outer perimeter (P_o) and area (A_o) (P_o is the perimeter of outermost adventitial border and A_o is the area enclosed by P_o). In cases in which the airway was contiguous to an adjacent vessel, an imaginary line was drawn between the two structures to estimate P_o. Airways in which over one third of P_o was contiguous to an adjacent vessel were omitted from the analysis. The inner wall area (WA_i), adventitial wall area (WA_o), and total wall area (WA_t) were calculated as: WA_i = A_mo  A_i, WA_o = A_o  A_mo, WA_t = A_o  A_i.

View larger version (33K): [in this window] [in a new window]   Figure 1.   The diagram of cross-sectioned membranous airway. DL and DS indicate long diameter and short diameters (see text). Pi, Pbm, Pmo, Po stand for internal perimeter, basement membrane perimeter, outer smooth muscle perimeter and adventitial perimeter respectively (see text).

Percent smooth muscle shortening (PMS) was calculated as: PMS = (P*_mo  P_mo)/P*_mo ×100, where P*_mo is the outer smooth muscle perimeter of a theoretical fully dilated airway and calculated as: P*_mo = [4(WA_i + A*_i)]^1/2 where A*_i is the theoretical fully dilated internal area, which was calculated based on the assumptions that P_i does not change and that P_i conforms to a perfect circle in the relaxed and dilated state using the equation: A*_i = P_i²/4 .

The number of eosinophils in each compartment of the airway wall was counted and expressed as density (number of cells/wall area). For identification of eosinophils, a ×100 objective lens was used.

To compare airways of different size, we used P_bm since the basement membrane was easier to trace than P_i, especially in bronchoconstricted airways where mucosal folds were tightly approximated. For calculating WA_i, we subtracted A_i from A_mo since A_i is relatively free from measurement error compared with P_i, even in the bronchoconstricted airways.

Statistical Analysis

All the data were expressed as mean ± SD. Ptr, BP, and HR at baseline and during vagal stimulation were compared using a two-tailed paired t test. Increase in Ptr before and during vagal stimulation at first and second stimulations in preliminary study was compared using a two-tailed paired t test. The frequency distributions of P_bm were compared between stimulated and unstimulated lungs using the Kolmogorov-Smirnov test. In the comparison of the wall compartments (WA_i, WA_o, and WA_t) between stimulated and unstimulated lungs, the square root of each wall area was used to linearize the relationship to P_bm. The slope and intercept of the relationships between P_bm and square root of each wall area for stimulated and unstimulated airways were compared using a two-tailed paired t test. In Protocol 1, one animal was eliminated from the comparison since only one airway was obtained from the stimulated lung. One-way ANOVA with blocking on animals was used to compare the eosinophil density in the stimulated (right lung) and unstimulated (left lung) airways. One-way ANOVA with blocking on animals was also used to compare the eosinophil density in the airways between Protocols 1 and 2. The random effects regression analysis (14) was used for the relationships between PMS and eosinophil density by pooling data and by taking the effect of electrical stimulation to the airways into account.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Protocol 1

In the preliminary study, baseline mean BP and HR for Protocol 1 were 37.2 ± 5.6 mm Hg and 264.2 ± 60.2 beat/min, respectively, and were significantly decreased during 2 min of unilateral vagal stimulation (33.2 ± 10.2 mm Hg and 155.7 ± 36.0 beat/min; p < 0.05).

The maximal Ptr was significantly increased after unilateral vagal stimulation compared with baseline condition (baseline, 25.5 ± 4.3 cm H₂O; during stimulation, 34.7 ± 2.8 cm H₂O; p < 0.05). In the morphometric measurements, 102 membranous airways were selected from eight animals (stimulated, n = 50; unstimulated, n = 52). The cumulative frequency distribution of P_bm was not different between the stimulated and unstimulated airways (Figure 2a), indicating that the sizes of airways that were selected for the analysis were comparable between stimulated and unstimulated lung. To compare the area of each airway compartment between the stimulated and unstimulated airways, the P_bm versus square root of wall area relationships were compared. There were no significant differences in the P_bm versus square root of wall area relationships of each compartment between the stimulated and unstimulated airways, indicating that the area of each airway compartment was comparable between stimulated and unstimulated airways over the range of airway sizes (Figure 3a). PMS in the stimulated airways (46.0 ± 16.1%) was significantly greater than that in the unstimulated airways (16.0 ± 14.5%), indicating that unilateral vagal stimulation created unilateral bronchoconstriction. Representative sections obtained from stimulated (a) and unstimulated (b) airways of same animal at low magnification are shown in Figure 4. A larger number of eosinophils was observed between epithelium and smooth muscle in the stimulated and constricted airways (see inset). The eosinophil density in the inner wall of the stimulated airways (204.2 ± 112.2/mm²) was significantly greater than that of the unstimulated airways (101.1 ± 83.8/mm²) (Figure 5a). The eosinophil density in the adventitia was significantly less in the stimulated (566.1 ± 472.6/mm²) than in the unstimulated (857.7 ± 609.2/mm²) airways (Figure 5a). There were no significant differences in eosinophil density in the total wall between stimulated (328.7 ± 186.6/mm²) and unstimulated (351.0 ± 206.3/mm²) airways (Figure 5a). PMS and eosinophil density were significantly and positively related in the inner wall (Figure 6a) (r = 0.46, p < 0.05) and significantly and negatively related in the adventitia (Figure 6b) (r = 0.40, p < 0.05).

View larger version (13K): [in this window] [in a new window]   Figure 2.   The cumulative frequency distribution of Pbm of stimulated (solid line) and unstimulated (broken line) airways. There were no significant differences between constricted and unconstricted airways for both protocol 1 (a) and protocol 2 (b) (see text).

View larger version (27K): [in this window] [in a new window]   Figure 3.   The relationship between Pbm and the square root of wall area for each compartment (WAi, WAo and WAt). Closed and open circles indicate stimulated and unstimulated airways, respectively. (a) and (b) indicate protocol 1 and protocol 2, respectively. Solid and broken lines indicate the regression lines for these relationships for stimulated and unstimulated airways from pooled data. There were no significant differences in the Pbm versus square root of wall area relationships for each airway wall compartment between stimulated and unstimulated airways in both experiments (a and b).

View larger version (116K): [in this window] [in a new window]   Figure 4.   Representative histology of stimulated (a) and unstimulated (b) airways of similar size (Pbm = 1.709 mm and 1.418 mm, respectively, original magnification ×200). Ep, M and Ad indicate epithelium, smooth muscle and adventitia, respectively. Black bars indicate 50 µm. Eosinophils are stained red using Hansel's staining. Many eosinophils are observed mainly between the epithelium and smooth muscle (a). Arrow indicates the area of inset which shows the accumulation of eosinophils (black bar indicates 10 micron, original magnification ×500).

View larger version (48K): [in this window] [in a new window]   Figure 5.   Eosinophil density in the inner wall (Eo/WAi) (top panels), adventitia (Eo/WAo) (middle panels) and total wall (Eo/WAt) (bottom panels) is compared between stimulated and unstimulated airways for both Protocol 1 (a) and Protocol 2 (b). The density of eosinophils was significantly greater in the inner wall and significantly less in the adventitia of stimulated than unstimulated airways in Protocol 1 (a) (*p < 0.05). There was no significant difference in the density of eosinophils in the total wall between stimulated and unstimulated airways in Protocol 1 (a). There was no significant difference in the density of eosinophils in any of the airway compartments between stimulated and unstimulated airways in Protocol 2 (b). The eosinophil density in the total wall was significantly larger in Protocol 1 than in Protocol 2.

View larger version (20K): [in this window] [in a new window]   Figure 6.   PMS versus eosinophil density relationship in the inner wall (a) and adventitia (b). Closed and open circles indicate stimulated and unstimulated airways. There was a significantly positive relationship in the inner wall (r = 0.455, p < 0.05) and a negative relationship in the outer wall (r = 0.4, p < 0.005).

Protocol 2

In the preliminary study, there were no significant differences in the increase in Ptr between first and second stimulations (5.23 ± 1.07 and 5.83 ± 0.65 cm H₂O, respectively), indicating that degree of bronchoconstriction was comparable between first and second stimulations.

Atropine treatment completely abolished the increase in Ptr during unilateral vagal stimulation (baseline, 23.3 ± 2.08 cm H₂O; after stimulation, 23.5 ± 3.45 cm H₂O). For the morphometric measurements, 88 membranous airways from eight animals were selected for analysis (stimulated, n = 46; unstimulated, n = 42). The cumulative frequency distribution of P_bm was not different between the stimulated and unstimulated airways, indicating that the sizes of selected airways were comparable between stimulated and unstimulated lung (Figure 2b). There were no significant differences in the P_bm versus square root of wall area relationships for each compartment between the stimulated and unstimulated airways, indicating that the area of each airway compartment was comparable between stimulated and unstimulated airways over the range of airway sizes (Figure 3b). There were no significant differences in PMS between the stimulated airways and unstimulated airways (13.2 ± 12.2 and 12.0 ± 12.4%, respectively). There were no significant differences in eosinophil density in the each airway wall compartment between stimulated and unstimulated airways (WA_i, 90.3 ± 77.2/mm² and 91.5 ± 85.5/mm²; WA_o, 612.9 ± 543.5/mm² and 635.6 ± 525.7/mm²; WA_t, 235 ± 136.7/ mm² and 252.6 ± 172.2/mm²) (Figure 5b).

Comparison of Eosinophil Density between Protocols 1 and 2

There was a significant increase in eosinophil density in the inner wall of stimulated airways in Protocol 1 compared with that of stimulated and unstimulated airways in Protocol 2 (Figure 5, top panels). There was no significant difference in the eosinophil density in the inner wall of unstimulated airways in Protocol 1 compared with that of stimulated and unstimulated airways in Protocol 2 (Figure 5, top panels). Eosinophil density in the adventitia of unstimulated airways in Protocol 1 was significantly greater than that of stimulated and unstimulated airways in Protocol 2 (Figure 5, middle panels). Eosinophil density in the total wall in Protocol 1 (both stimulated and unstimulated) was significantly greater than that in Protocol 2 (Figure 5, bottom panels).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study, we used electrical stimulation of the right vagus to create unilateral bronchoconstriction and compared the eosinophil accumulation between the stimulated and contralateral unstimulated airways. Although there was some overlap, PMS was significantly greater in the stimulated than in the unstimulated airways (46.0 ± 16.1 and 16.0 ± 14.5%, respectively, see also Figure 6), suggesting that the unilateral vagal stimulation effectively created unilateral bronchoconstriction in the right lung. Using this model, we investigated the accumulation of eosinophils in the constricted airways by comparing them with unconstricted airways as a control in the same animal. To compare the accumulation of eosinophils in the airway wall between constricted and unconstricted airways, we used eosinophil density rather than actual counts since the number of eosinophils is bound to be influenced by the variation in the wall area that is observed in airways of similar size (15). Results show that eosinophil density was significantly increased in the inner wall and was significantly decreased in the adventitia of constricted airways compared with unconstricted airways (Figure 5a). Because there was no significant difference in the area of each airway wall compartment between constricted and unconstricted airways in both protocols, eosinophil density in the inner wall is attributed to the accumulation of eosinophils to the inner wall. In Protocol 2, vagal stimulation was applied twice with a 30-min interval, and there were no significant changes in eosinophil density in both inner wall and adventitia after the second stimulation under atropine treatment. These observations suggest that increased eosinophil density during the first electrical stimulation must have returned to baseline before the second stimulation and that such a quick change in eosinophil density is probably due to the change in cell number in the bronchial vasculature.

Vagal stimulation releases several chemical transmitters, including acetylcholine, neurokinins, VIP, and nitric oxide (16). Because substance P can attract eosinophils (20), the accumulation of eosinophils in the inner wall during vagal nerve stimulation in Protocol 1 could have been the result of chemotactic activity. The accumulation of eosinophils in the inner wall in this study, however, was completely abolished by pretreatment with atropine. This observation suggests that e-NANC activation was not the main reason for the accumulation of eosinophils during vagal stimulation. Another possibility for the accumulation of the eosinophils during vagal stimulation is the activation of other chemotactic factors or adhesion molecules of eosinophils caused by acetylcholine release, although this is unlikely.

Studies from our laboratory have shown that PMN are delayed with respect to red blood cells during a single passage through the pulmonary circulation (21). Doerschuk and coworkers (22) have presented evidence that suggests that this retention is due to mechanical factors. Markos and colleagues (23) have shown that application of positive end-expiratory pressure, which is expected to compress alveolar vessels, enhanced the pulmonary retention of PMN, and that the effect was immediately reversed on removal of the positive end- expiratory pressure, again suggesting that mechanical factors are important for margination of PBM in pulmonary circulation. Because the mucosal membrane internal to the smooth muscle layer has to deform and be thrown into folds during bronchoconstriction, it is possible that mechanical retardation of eosinophils could occur as a result of geometric deformation of the bronchial vascular bed.

Regional blood flow could also be a factor affecting the accumulation of leukocytes in the capillary network (24). BP and HR decreased during vagal stimulation in the animals without atropine pretreatment, indicating that cardiac output probably decreased during this period. Indeed, the eosinophil density in the total wall in Protocol 1 was significantly greater than that in Protocol 2 in which BP and HR were probably unchanged because of atropine pretreatment in the latter (Figure 5). In Protocol 1, we created constricted and unconstricted airways in the same animal to minimize the effect of cardiac output on the accumulation of eosinophils, although local blood flow could be different between stimulated and unstimulated sides. Because blood flow to the airway mucosa rather increases during vagal stimulation (25), density of eosinophils should decrease in the inner wall; this was not the case in our results.

Using histologic specimens, Miller (26) postulated that the bronchial arteries enter the airway wall from the adventitial side and penetrate into the mucosal membrane internal to smooth muscle layer to create a capillary network. He also postulated that the venous channels penetrate the smooth muscle layer and form a venous plexus in the adventitia (i.e., venous plexus is in series with the capillary network). During bronchoconstriction, the inner wall is thrown into folds and the capillary network in the inner wall is likely to be deformed because of the increase in interstitial pressure. It is possible that eosinophils are retarded in the deformed capillary of the inner wall. Indeed there was a significant and positive relationship between the magnitude of the deformation (represented by PMS) and the eosinophil density in the inner wall (Figure 6a). If eosinophils are retarded in the capillary network in the inner wall and the blood stream subsequently penetrates into the venous plexus in the adventitia as Miller postulated, the density of eosinophils in the adventitia should decrease during bronchoconstriction. Indeed, the eosinophil density in the adventitia was decreased (Figure 5a). The relationship between the magnitude of airway wall deformation (i.e., PMS) and eosinophil density in the adventitia was also significantly negative (Figure 6b), suggesting that the retardation of eosinophils in the inner wall could cause the decreased density of eosinophils in the adventitia. Our results, therefore, indirectly support the anatomic arrangement of the bronchial circulation in the airway wall proposed by Miller.

In summary, we have shown the accumulation of eosinophils in the inner wall of the constricted airways in guinea pigs using a unilateral vagal stimulation model. We speculate that mechanical factors could cause retention of eosinophils in the deformed airway wall during bronchoconstriction.