Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Heightened responses to varying concentrations of bronchoconstrictors such as methacholine (airway hyperresponsiveness) and airway obstruction (AO) are characteristic of asthma and common in other clinical states, including chronic obstructive pulmonary disease (COPD), lung transplantation, and after airway inflammation induced by infection or chemical agents. Factors postulated to influence airway physiology include airway caliber, inflammation and alterations in airway mechanics, autonomic regulation, and epithelial integrity (1). The extent to which various types of structural changes contribute to these physiologic abnormalities, however, is poorly understood.

Airway inflammation and cytokine dysregulation are felt to be central features of the pathogenesis of asthma and asthmatic airway hyperresponsiveness (AHR). However, the effector functions and physiologic consequences of many cytokines in the airway can frequently only be speculated upon. To characterize the in vivo effector functions of interleukin (IL)-6 and IL-11 in the airway we generated transgenic mice that overexpressed these cytokines under the control of the Clara cell 10-kD protein promoter (CC10) (4, 5). IL-6 and IL-11 were chosen for comparison because they are both members of the IL-6-type cytokine family and have similar three-dimensional structures (6). They also have overlapping biologic effector profiles that can be, at least partially, attributed to their shared use of the gp-130 molecule as the signal transducing  subunits of their multimeric receptor complexes (7, 8). As previously reported, the phenotypes of these animals have a number of similarities, including the accumulation of lymphoid nodules in peribronchiolar regions and airway wall thickening (4, 5). Our previous studies also demonstrated that the IL-11 mice have emphysema-like airspace enlargement that is the result of the inhibition of septation during acinar development (9) and airway remodeling with subepithelial fibrosis (5). Surprisingly, the airway physiology of these animals differed markedly. The animals that express IL-6 manifested normal baseline airways resistance and were hyporesponsive, requiring a 5- to 10-fold higher dose of methacholine than did their littermates to double their airflow resistance. In contrast, baseline airway resistance (Raw) in mice expressing IL-11 was increased 3-fold over that of normal littermates. These animals also manifest AHR, doubling their Raw at only one-twentieth the dose of methacholine required by the transgene negative littermate control mice.

To gain insight into the pathologic processes that contribute to the different physiologic profiles of the IL-6 and IL-11 transgenic mice we compared the histologic, morphometric, and physiologic features of these animals. These studies demonstrate that IL-6 overexpression also causes alveolar airspace enlargement and subepithelial fibrosis and that similar degrees of airspace enlargement are seen in the IL-6 and IL-11 transgenic animals. They also demonstrate that there was a significant increase in the caliber of the remodeled bronchioles in the hyporesponsive IL-6 animals. In contrast, remodeling occurred in airways with normal external and luminal dimensions in the hyperresponsive and obstructed IL-11 transgenic mice.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Generation of IL-11 and IL-6 Transgenic Mice

The techniques that were used to generate the mice that overexpress IL-6 (CC10-IL-6) and IL-11 (CC10-IL-11) have been previously described (4, 5). Both used CC10 (a generous gift of Drs. Jeff Whittsett and Barry Stripp, University of Cincinnati, Cincinnati, OH) to target the cytokines to the lung. Appropriate constructs were prepared, junctions were confirmed by sequencing, the constructs were linearized and purified, and transgenic animals were generated via microinjection using standard techniques. Multiple independent lines expressing each cytokine have been propagated. The presence or absence of the transgene in the offspring from these founder mice was determined by Southern or polymerase chain reaction analysis of DNA isolated from tail clippings, and the concentrations of IL-6 and IL-11 protein in airway bronchoalveolar lavage fluid (BALF) were assessed by enzyme-linked immunosorbent assay. Mice with similar concentrations of BALF IL-6 and IL-11 were used for the comparisons in these studies.

Histologic Analysis

Transgenic mice and littermate control mice were sacrificed at 4 to 8 wk of age. The lungs were fixed to a set pressure by intratracheal instillation of 3% gluteraldehyde from a reservoir 25 cm above the mouse, and both lungs were completely embedded and sections prepared. Routine hematoxylin and eosin and Masson's trichrome stains were employed.

Lung Volume Measurements

Transgene () and transgene (+) mice were killed at 6 to 8 wk of age, and their lungs and heart were removed en bloc and fixed at 25 cm as described previously. The volume of the lungs and heart was then measured by quantitating the volume of water displaced by the inflated organs in a fluid-filled graduated cylinder. The volume of the heart was then assessed separately and subtracted from the lung-heart value.

Morphometric Evaluation

Morphometric evaluation of airspace enlargement was carried out as previously described (9) using digitized video images acquired at ×50 magnification using an Olympus BH2 microscope (Olympus, Tokyo, Japan) and a Hitachi KP160 CCD camera (Hitachi Instruments, San Jose, CA). The video output of the camera was sent to a Perceptics imaging microscope workstation (Perceptics Corporation, Knoxville, TN) with a model 9201 image processor interfaced to an Apple Macintosh IIfx computer. For the alveolar measurements, both lungs were fixed, embedded en bloc, sectioned in the sagittal plane, stained, and scanned in their entirety by a technician who acquired all microscopic fields that lacked major bronchovascular structures. The area of each field analyzed was 62,735 µm² and an average of 15 fields per slide was captured. Images were analyzed using Biovision Image Analysis Software (Perceptics Corporation). Two independent measurements were recorded on each image: the average of the chord lengths (distance between airspace walls) and the length of the air-tissue interface. The latter, when divided by the area of the image, is the equivalent of the surface area of airspace wall per unit lung volume.

Bronchiolar measurements were made at ×400 magnification using an eyepiece reticle that was calibrated against a stage micrometer. Bronchioles less than 250 µm in diameter that presented a smooth circular or oval profile were selected, and measurements were recorded to the nearest half division on the reticle (1.25 µm). As illustrated in Figure 1, the measurements were made across the short axis of the elliptical profiles at locations where the cell borders appeared sharp to minimize the influence of tangential sectioning. Measurements included the lumen diameter ("d" in Figure 1), the diameter of the airway between the outermost layers of smooth muscle ("c" in Figure 1), and the thickness of the bronchiolar wall from the base of the columnar epithelium to the outer limit of the adventitia ("e" in Figure 1). On a subset of airways (airways that did not share an extensive length of their adventitia with arteries), both the short ("a" in Figure 1) and long ("b" in Figure 1) axes of the elliptical profiles were measured, and the number of alveolar walls attached to the bronchiole were recorded.

View larger version (28K): [in this window] [in a new window]   Figure 1.   Schematic representation of variables for airway morphometry. a = external (transverse) diameter; b = longitudinal diameter; c = distance between smooth muscle layers; d = lumen diameter; e = wall thickness; f = thickness of mucosa and submucosa calculated from measurements c and d. The longitudinal diameter of the airway was only used to calculate the outside perimeter of the airway (see MATERIALS AND METHODS).

From these measurements, two additional endpoints were calculated: thickness of the inner airway wall, including epithelium, lamina propria, and smooth muscle ("f" in Figure 1), was calculated as one-half the difference between the lumen diameter and the diameter across the bronchiole measured between smooth muscle layers,

The perimeter of each airway was calculated using the approximate formula for the circumference of an ellipse,

where a' and b' are one-half of the short (a) and long (b) axes of the ellipse. Alveolar attachments per millimeter were the quotient of the number of alveolar walls attached to a bronchiole divided by the length of the perimeter of that bronchiole. The mean ± standard deviation (SD) number of bronchioles measured per mouse to determine bronchiolar diameters and wall thickness was 11.1 ± 2.5; the number for perimeter and alveolar attachments was 8.9 ± 1.9.

For statistics, the values of each variable for a given mouse were averaged to give a single value for that mouse. Using Instat 2.03 software for the Macintosh, the means and standard deviations for all mice in a group were calculated. Group differences were evaluated by analysis of variance and when the analysis of variance indicated significant differences, the differences between individual groups were tested by Student's t test with Bonferroni's correction for multiple comparisons.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Inspection and morphometric evaluation of sections from the CC10-IL-6 and CC10-IL-11 transgenic animals demonstrated alveolar and airway abnormalities when compared with littermate control mice.

Alveolar Alterations

As previously described, the airspaces of the CC10-IL-11 mice were markedly enlarged compared with wild-type littermate control mice (9). When the CC10-IL-6 animals were fixed and evaluated in an identical fashion similar airspace enlargement was noted (Figure 2). These emphysematous changes were readily confirmed with the morphometric evaluations. As seen in Table 1, the airspaces in the CC10-IL-6 and CC10-IL-11 mice were at least twice as large as the airspaces from littermate control mice (P < 0.001 for CC10-IL-6 and CC10-IL-11 mice compared with control mice). Similarly, both cytokine transgenic strains had diminished surface/volume ratios and a density of bronchiolar alveolar wall attachments that was half of that seen in control mice (P < 0.001 for CC10-IL-6 and CC10- IL-11 mice for both parameters when compared with littermate controls) (Table 1). In accord with these findings, the volumes of lungs from the IL-6 and IL-11 mice were approximately twice that of the control mice, and their static compliance values at this inflation pressure were similarly increased (Table 2) (P < 0.05 for CC10-IL-6 and CC10- IL-11 mice for both parameters versus littermate controls). In contrast to the readily apparent differences in the airway physiology of the CC10-IL-6 and CC10-IL-11 mice, differences in the alveolar parameters of these animals were not readily apparent. The assessments of chord length, surface/volume ratio, and number of alveolar attachments per unit bronchiolar perimeter, lung volume, and static compliance were similar in the CC10-IL-6 and the CC10- IL-11 animals (Table 1). This suggests that the airspace alterations in the CC10-IL-6 and CC10-IL-11 animals do not, by themselves, determine the differences in airways resistance or methacholine responsiveness that were noted.

View larger version (102K): [in this window] [in a new window]   Figure 2.   Alveolar abnormalities in transgenic mice. Hematoxylin and eosin stain of lung tissue from transgene () (normal, top panel), IL-6 transgene (+) (middle panel), and IL-11 transgene (+) (bottom panel) mice. (All photographs are at ×10 magnification.)

Airway Alterations

A variety of airway abnormalities were noted in the transgene mice.

Leukocyte abnormalities. Both the CC10-IL-6 and CC10- IL-11 animals had nodular collections of peribronchiolar mononuclear cells. Similar collections were not noted in the littermate control animals (Figure 3). Differences in the location, appearance, or frequency of these structures between the CC10-IL-6 and CC10-IL-11 animals were not readily apparent (data not shown).

View larger version (88K): [in this window] [in a new window]   Figure 3.   Nodular collections of mononuclear cells in transgenic mice. Hematoxylin and eosin stains demonstrating the presence of nodular collections of mononuclear cells in IL-6 (middle panel) and IL-11 (bottom panel) transgenic mice and the lack of these collections in wild-type mice (normal, top panel). (All photographs are at ×10 magnification.)

Wall abnormalities. On light microscopic examination, the walls of the bronchioles from the CC10-IL-6 and CC10-IL-11 mice were thickened when compared with control mice. This was partially the result of an increase in cellularity. The control bronchioles were so thin that the structure of the walls was difficult to resolve and only a single layer of elongated nuclei was observed. In contrast, three to four layers of nuclei were commonly noted in the walls of the bronchioles from both of the transgenic strains (Figure 4). Collagen deposition also contributed to the bronchiolar changes because the enhanced accumulation of subepithelial and adventitial collagen was readily apparent in trichrome stains in both transgenic strains compared with littermate controls (Figure 5).

View larger version (79K): [in this window] [in a new window]   Figure 4.   Histology of bronchioles of wild-type and transgenic mice. Hematoxylin and eosin stains demonstrating the bronchiolar structures of transgene () mice (normal, top panel), IL-6 transgene (+) mice (middle panel), and IL-11 transgene (+) mice (bottom panel). (All photographs are at ×50 magnification.)

View larger version (93K): [in this window] [in a new window]   Figure 5.   Collagen accumulation in wild-type and transgenic bronchioles. Trichrome evaluation of bronchiolar structures of transgene () (normal, top panel), IL-6 transgene (+) (middle panel), and IL-11 transgene (+) (bottom panel) mice. Collagen stains blue on this evaluation. (All photographs are at ×50 magnification.)

View larger version (14K): [in this window] [in a new window]   Figure 6.   Size distribution of all measured bronchioles in transgene () and transgene (+) mice. (Open squares, control; closed squares, IL-11; inverted triangles, IL-6.) The data illustrate the cumulative percentage of bronchioles of less than or equal to a given size.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Physiologic abnormalities contribute to the morbidity and mortality of disorders such as COPD and asthma. AHR, an exaggerated bronchospastic response to nonspecific stimulants such as methacholine, and AO are cornerstones of this physiologic dysfunction. It has been assumed that the structural abnormalities that are well described in these disorders contribute to the pathogenesis of these physiologic abnormalities. The types of structural changes that cause AHR and AO are, however, poorly defined. To address this issue, we took advantage of transgenic models prepared in our laboratory in which the overexpression of closely related cytokines under the influence of the same promoter resulted in mice that have normal baseline airway resistance and a diminished bronchospastic response to methacholine (CC10-IL-6 mice) and mice with airflow obstruction and a hyperresponsive reaction to methacholine (CC10-IL-11 mice). In these studies we demonstrated that the light microscopic phenotype of these animals is similar in several respects with both transgenic strains manifesting peribronchiolar lymphocytic nodules, airspace enlargement, and airway wall thickening with subepithelial fibrosis. Importantly, we also applied morphometric approaches to these transgenic models. These quantitative structural evaluations revealed differences between the two transgenic strains that may help clarify the mechanism(s) of these physiologic differences.

To understand the physiologic consequences of airway remodeling, several investigators have carefully analyzed the structural features of human and experimentally manipulated airways that are associated with altered bronchospastic responses to nonspecific agonists (10). Wiggs and colleagues (13, 14) developed computer models describing airway responses to bronchoconstrictors. In their model, airway narrowing brought about by smooth muscle shortening is opposed by the elastic recoil of the lung and collagen, hyaluronic acid, and versican in and around the airway. In this model, the degree of luminal effacement in response to a given stimulus for smooth muscle contraction is augmented by an increase in the volume of the inner airway wall (tissue between the luminal surface and smooth muscle layer), adventital thickening, increased smooth muscle mass (with preserved or heightened contractility), and/or a decrease in parenchymal elastic recoil (11, 20). The effects of airway thickening, muscle mass, and changes in elastic recoil were more than additive (14, 15), and maximal airway obstruction occurred when the increase in airway thickness was out of proportion to airway size (17). In studies using human lungs, lumen size can be difficult to assess, and airways can have a crenated outline because lung expansion is difficult to control and epithelial sloughing can occur. This led them to use the airway epithelial basement membrane perimeter as an index of airway size (10, 17, 19). Our studies are the first to test these hypotheses in an in vivo genetically controlled system. In these studies, we used direct measurements of lumen dimensions because our glutaraldehyde fixation technique allowed lung expansion to be readily controlled and avoided epithelial sloughing. These studies revealed three impressive differences between the IL-6 and IL-11 animals that support the proposals of Pare, Wiggs, and colleagues. First, the bronchiolar wall thickness of the CC10-IL-11 transgenics was greater than in the IL-6 transgenics. Second, the bronchioles in the CC10-IL-6 animals were significantly larger than in littermate control animals or in the CC10-IL-11 animals. As a result, wall thickening in the CC10-IL-6 animals was in proportion to the enhanced airway size as shown by the normal values obtained when airway wall thickness was normalized to airway external diameter. In contrast, the thickening in the IL-11 mice was out of proportion to airway size with significantly increased values when wall thickness was normalized in a similar fashion. Lastly, the IL-11 animals had significantly enhanced inner wall thickness when compared with littermate control animals, whereas the IL-6 animals did not. It is thus easy to see how the enhanced size of the airways of the IL-6 mice might serve to diminish the airway narrowing caused by agonist-induced smooth muscle contraction, whereas the combination of airway wall thickening out of proportion to airway size and submucosal thickening would augment the resistance to airflow induced by a similar smooth muscle response in the CC10-IL-11 mice.

Computer modeling and other investigations of COPD and other airways disorders suggest that AHR and AO can be caused by parenchymal destruction, which decreases parenchymal elastic recoil and radially directed airway tethering (17, 19). The loss of tethering forces would be even more profound in the setting of airway wall thickening, which is postulated to uncouple airway smooth muscle from the surrounding parenchyma (17, 19, 20). In addition to bronchiolar wall thickening, alveolar enlargement and enhanced compliance are features of both the IL-6 and IL-11 transgenic mice. Surprisingly, our studies demonstrated that IL-6 and IL-11 mice, which have remarkably different physiologic profiles, have comparable degrees of airspace enlargement. Thus, alveolar enlargement and compliance alterations do not always cause AO or AHR. Although these changes could contribute to the AHR and AO seen in the CC10-IL-11 mice, they cannot, by themselves, fully account for the physiologic profile of these animals. In addition, if airway enlargement and compliance alterations contribute to the physiologic profile of the CC10-IL-6 mice, the tendency of these alterations to induce AHR and AO must be overcome by the opposing effects of the bronchiolar enlargement that occurs at the same time (see subsequent paragraph). Thus, in these murine modeling systems, significant airspace enlargement is less important than airway size in determining responsiveness to methacholine challenge.

Poiseuille's law states that the resistance to laminar flow in a tube (an approximation of an airway) varies inversely with the fourth power of the radius. Hence, it is easy to see how the enlarged luminal diameter could account for the unusually low airflow resistance after bronchoconstriction in the CC10-IL-6 mice. It is also easy to see how the bronchiolar enlargement in the IL-6 mice could override and the normal airway caliber in the CC10- IL-11 mice could augment the AHR- and AO-generating effects of the bronchial wall thickening and "emphysema" in these animals. Thus, the increased caliber of the bronchioles of the CC10-IL-6 mice relative to the CC10-IL-11 mice provides a plausible explanation for their different physiologic phenotypes. These studies therefore demonstrate that airway remodeling is not inevitably associated with AO or AHR in this murine system. The caliber of the airways in which the remodeling occurs also determines the physiologic outcome. In human studies, an inverse relationship between airway caliber and AHR has also been reported (22). However, airway caliber, in these studies, was inferred from measurements of baseline expiratory flow. The present studies confirm the importance of airway size as a determinant of AHR and AO in a system in which airway size can be measured directly.

The mechanism of airway enlargement in the IL-6 animals is not clear. Because the CC10 promoter is known to initiate gene expression during crucial periods of lung development (9), we speculate that this IL-6-induced alteration in airway size may be developmentally mediated. This speculation is in accord with studies from this laboratory and others that demonstrate that IL-11 can alter lung alveolar development (9), that early life events can lead to airway hypoplasia and physiologic dysregulation (25), and that developmental processes play an important role in the pathogenesis of asthma (26). They also raise the possibility that early life stimuli that induce airway remodeling responses generate AHR only when this occurs in airways with normal or compromised luminal dimensions and that measurements of airway wall thickness and airway size are important parameters that need to be evaluated in future studies attempting to understand the pathogenesis of human airway disorders such as asthma and COPD.

The studies described in this report provide evidence of the importance of structural alterations in the pathogenesis of AHR and AO. They do not, however, rule out other structural, inflammatory, neural, or chemical processes that might also contribute to the different physiologic profiles of the control and transgenic animals. Such processes include abnormalities in smooth muscle, possible alterations in neural structures and muscle contractility, and charge effects of IL-11 in these animals. The muscle changes may be particularly important because increases in smooth muscle mass can contribute in a major way to AHR (17, 19) and the CC10-IL-11 transgenic animals manifest increases in airway -smooth muscle actin staining cells that are myocytes and myofibroblasts (5). If similar alterations are seen in the CC10-IL-6 animals, quantification will need to be undertaken. This will be difficult, however, because myofibroblasts and myocytes cannot be easily differentiated, and the extremely thin muscle bundles of the control lungs cannot be easily resolved by light microscopy. IL-6 has also been shown to inhibit vascular smooth muscle contractility (27). In addition, IL-11 (28), but not IL-6 (29), is extremely cationic (pI 11.7) and cationic proteins can induce AHR by charge interactions with the airway wall (30, 31). Lastly, IL-6 and IL-11 could also generate their respective phenotypic profiles via their ability to induce or not induce cholinergic differentiation and tachykinin production by neural structures in respiratory airways (32). Additional investigations will be required to evaluate the contributions of each of these processes.

In summary, we compared the histologic and morphologic features of transgenic mice that overexpress IL-6 and IL-11 and differ in their levels of baseline airway resistance and responsiveness to methacholine. The lungs of both groups of transgenic mice showed similar peribronchiolar nodules, alveolar enlargement, and airway remodeling with subepithelial fibrosis. The severity of the alveolar airspace enlargement was similar in both groups of transgenic mice and therefore could not account for their different physiologic profiles. Morphometry demonstrated that the wall thickening was more robust in the IL-11 transgenic mice and airway caliber was increased in the IL-6 transgenic mice. Thus, normal airflow and airway hyporesponsiveness were noted when the remodeling response occurred in enlarged airways, and AO and AHR were noted when remodeling occurred in normal-sized airways and was associated with wall thickening out of proportion to airway caliber. These studies demonstrate that airway remodeling and alveolar enlargement do not uniformly lead to AHR and AO. The severity of the remodeling response and the caliber of the airway are important additional determinants of whether these responses induce or do not induce AO and/or hyperresponsiveness to nonspecific agonists such as methacholine.