Histomorphometric study of the pulmonary response of guinea pigs to chronic cotton dust inhalation

Histomorphometric

Study of the Pulmonary Response of Guinea Pigs to Chronic Cotton Dust Inhalation’

P. A. COULOMBE,’

P. R. FILION,

AND M. G. COTTA

Dkpartement de pharmacologic. Fault& de mkdecine. Universitk de Montrkal. C.P. 6128, succursale “‘A ‘*,Montrt?al, Qukbec. Canada, H3C 3J7

Received January 31. 1986; accepted May 19. I986 Histomorphometric Study of the Pulmonary Response of Guinea Pigs to Chronic Cotton Dust Inhalation. COULOMBE, P. A., FILION, P. R., AND C&T& M. G. (1986). Tosicol. Appl. Pharmacol. 85, 437-449. In this study we describe, through stereological methods, the lung morphology following inhalation exposure of guinea pigs to 2 1 mg/m’ cotton dust (CD) for 1 year. Various stereological parameters were determined on semithin histological sections. through a multistage sampling approach, to study the reaction of the whole lung, alveolar parenchyma. and bronchioles to CD inhalation. After 1 year of exposure, the lung volume was increased. Two distinct patterns of lung response were identified among the exposed animals. In type I responders, most of the morphometric parameters measured to describe the alveolar parenchymal reaction were within control value limits (X f 2 SD). In type II responders, the volume density ( C-;)of the parenchymal zone was decreased, while the b;, mean thickness, and absolute volume of the alveolar septa were increased. These changes caused the surface density (S,) of alveolar epithelium to decrease, and an estimate of the percentage of alveolar septa remaining functional for gas exchange was also significantly lowered in these animals. In both types of responders, fifth to ninth orders of bronchioles had a raised wall to lumen ratio: the Vu and mean thickness ofthe bronchiolar epithelium were markedly increased, denoting hyperplastic changes. Thus, chronic exposure to cotton dust induced definite morphological changes on the peripheral conducting airways in most of the exposed animals, and induced pronounced changes at the alveolar level in 8 of 17 CD-exposed guinea pigs. 0 1986 Academic PWS, ITIC.

The daily exposure of industrial workers to cotton fibers for many years results in a progressive impairment of lung function which manifests itself through the clinical symptoms of byssinosis, including chest tightness, breathlessness, and cough. Numerous studies conducted in human byssinotic subjects have focused on the pulmonary ventilatory mechanics, which appeared altered in this occu’ This investigation was supported by Universitt de Montreal internal funds (CAFIR). ’ Holder of a graduate studentship from the Medical Research Council of Canada. ‘To whom all correspondence and reprint requests should be addressed.

pational disease (Rylander et al., 1983; Schachter et al., 1984). However, the pulmonary morphology following chronic cotton dust (CD) exposure needs be characterized (Pratt et al., 1980; Schachter et al., 1984). Recently, the guinea pig was proposed as a useful animal model to document the acute and chronic effects of CD exposure on the respiratory function (Ellakani et af., 1984, 1985). These investigators were able to reproduce in this rodent the clinical course of respiratory symptoms associated with human byssinosis (Ellakani, 1985; Alarie et al., 1985; Ellakani et al., 1985). The lungs of these animals thus offer the opportunity of investigating the morphological aspects of the reaction to chronic CD challenge. 437

0041-008X/86

$3.00

Copyright 0 1986 by Academic Press, Inc. All rights of reproduction in any form reserved.

438

COULOMBE.

FILION,

We recently demonstrated that stereological methods could be successfully applied to plastic-embedded histological material, so as to document the morphological reaction of the pulmonary alveoli to a toxic challenge (Coulombe et al., 1985). In the present study, we further expanded this stereological approach by partially characterizing the lung morphology resulting from 1 year inhalation exposure of guinea pigs to CD. METHODS Source of the lungs. The CD-exposed lungs used in this study were obtained from the University of Pittsburgh. Details pertaining to the experimental procedure are given elsewhere (Ellakani, 1985: Ellakani et al.. 1985). Briefly, male English smooth-haired guinea pigs (HLA (FR)BR, Hilltop Lab Animals, Scottdale. Pa.) were divided into control and exposed groups of 20 animals each. The latter group was exposed 6 hr/day, 5 days/ week, for 52 weeks to a CD concentration of20.96 f 1.49 mg/m3. The mean CD particle size was 2.4 * 0.3 pm aerodynamic equivalent diameter. During the course of the study, the health status of the guinea pigs was monitored daily through measurement of body weight and gross clinical examination. The weight gain was similar in both groups until Day 200: the growth thereafter was slower in exposed than in control animals (Ellakani, 1985). Of the 34 animals used in the present study, none had any sign of spontaneous disease that could have influenced the pulmonary morphology. Eighteen hours following the 364th day of exposure the animals were anesthetized (Nembutal 60 mg/kg ip) and the lung-trachea complex was sectioned below the cricoid cartilage and dissected out of the thoracic cavity. The volume of the collapsed lungs was determined by water displacement prior to fixation. After cannulation and ligation of the trachea, the lung tissue was fixed by a gravity-assisted instillation (20 cm HI0 pressure) of 10% buffered Formalin for I hr. The cranial and caudal lobes of the right lung from 34 animals (17 controls, I7 exposed) were then carefully packed in formaldehyde and sent to the Universite de Montreal. Until the morphometric study was completed, lung samples were identified only by a randomization code.

Embedding and sectioning of the pulmonary tissues. Fixed lung lobes were cut along their cephalo-caudal axis into slices 2 mm thick, and the most external slices were discarded. Four of these slices from each animal (two per lobe) were further thinned, dehydrated in 50% ethanol. and embedded in glycol methacrylate (GMA) according to a method described elsewhere (Sims, 1974; Bemier ez al., 1984). After vacuum polymerization, semithin histo-

AND

COTE

logical sections (0.5-0.7 pm) were cut on a Sorvall JB4 microtome and stained with 0.X toluidine blue. In addition, one lung slice from each cranial lobe was embedded in paraffin, from which thick sections (5 pm) were stained by the Masson trichrome procedure for the qualitative study of the connective tissue distribution (Luna. 1968). Only the GMA-embedded material was used throughout the quantitative microscopic study. as it was more suitable for that purpose (Helander. 1983 ). Stereologicol methodology, The various compartments ofthe lung were quantitied using a multistage sampling, which consists of a cascade of several sampling stages with progressively increased magnifications. taking into consideration the different order of magnitude of the units which make up an organ (Cruz-Orive and Weibel. 198 I ). The characteristics of the test systems and sampling requirements of each of those sampling stages were experimentally determined using methods described by Weibel (1979). No attempt was made to correct for errors arising from section thickness (the Holmes effect; Holmes, 1921). since the compartments investigated were large in comparison with this thickness. Ana/.wis of‘ the ushole lun,c. We first determined at a low magnification the volume densities ( I;) of each of the lung histological compartments (airways, vessels, parenchyma) through a point-counting approach. with the whole lung as the reference space. The analysis was performed directly under an optical microscope (Swift M 1000-D), equipped with an ocular grid containing I2 I symmetric points. at a final magnification of lOOi.. The point count was registered manually. We analyzed five microscopic fields from each of four histological sections per animal, providing a total of 20 fields and 2420 points evaluated per animal. The interassay variability of this point-counting method was measured as less than 6”;‘. The absolute volume of each of those compartments could then be calculated, using the volume of the whole lung initially measured. :lna!,:ci.r o/‘lhe alveolar parenc~lr~wza. At a second level of sampling, the parameter I ‘,, (alveolar septum. parenchyma) was determined using the linear integration principle of Rosiwal (Weibel, 1979). This measure refers to the volumetric fraction ofthe lung parenchyma occupied by septal tissue. The parameter surface density (S,) (alveolar epithelium, parenchyma) was measured using an equation independently derived by Tomkeieff and Saltykov (Weibel. 1979). This measure refers to the surface area of alveolar epithelium expressed relative to unit volume ofparenchyma (units mm’/mm’). In both cases, the reference space was the lung parenchyma. We also studied the frequency distribution of the thickness ofthe alveolar septa by the line-intercept method. Measures of septal thickness within the range O-50 pm were distributed among 20 classes (class interval: 2.5 pm). Approximately 500 independent measures of this septal thickness were registered for each animal. From these measurements were calculated the arithmetic mean thickness of the

MORPHOMETRY

OF

COTTON

septa (T), the percentage of alveolar septa functional for gas exchanges (% Funct septum), and the parameters .d (slope) and B (ordinate) ofthe frequency distribution. The percentage of functional alveolar septa is a measure of the proportion of these structures that permits normal gas exchanges. On the basis of Weibel’s work ( 1973). it was assumed that normal oxygenation of the blood flowing into the alveolar capillary circuit could occur in septa of thickness less than five times the control value. as measured in I7 sham-exposed animals. It was also assumed in this study that the alveolo-capillary barrier reacts to adverse stimuli in a pattern very similar to that of the whole septum itself. The relationship between the accumulated frequencies and the septal thickness could be linearized by means of standard mathematical transformations (logit for the .Y axis. logarithmic for the I’ axis). The linear regression coefficient (2) calculated from this linearized relationship was greater than 0.90 and statistically significant in all control and exposed animals, Therefore, a slope and an ordinate could be calculated to further document the frequency distribution of the septal thickness following CD exposure. We randomly sampled the parenchyma from the cranial lobe by photomicrography. at a magnification of 700X (Zeiss Photomicroscope 11). Each photomicrograph was analyzed by projection onto a quadratic test grid of dimensions 18 X 23 cm, comprising seven lines of 23 cm length. The intercepts made by the tissue structure ofinterest with the test lines were electronically measured (Zeiss MOP III). Forty microscopic fields per animal were analyzed at a final magnification of2000X. thus providing a total sampled surface per animal of 0.432 mm’. The interassay variability of this method was less than 10%‘. Finally, the parameter N, (number of a cell of one type per unit surface area) was determined for the alveolar macrophages. type II pneumocytes, and intermediate pneumocytes by the same sampling procedure described above. In the pulmonary alveoli. type II pneumocytes maintain the integrity of the respiratory epithehum through its capability to differentiate into type I pneumocytes (Adamson and Bowden. 1974: Evans et nl , 1975). Along this differentiation process, a transient cell form called intermediate pneumocyte appears and is characterized in optical microscopy by a flattened and elongated cell shape as compared to type II pneumocytes, while still featuring the characteristic lamellar bodies. An index of alveolar epithelial repair could be calculated as the percentage of the total alveolar epithelial regenerative pool (type II + intermediate pneumocytes) which existed under the differentiating intermediate form in the alveoli. .-lnalysis ofthe distalairways. The sampled population in this study consisted of airways from the fifth to ninth orders of the respiratory tract, and thus was restricted to bronchioles. This classification was based on the comparison of the airway surface area and maximal diameter values with the data reported by Schreider and Hutchens

DUST-EXPOSED

LUNGS

439

(1980) for epoxy casting of the guinea pig respiratory tract. A systematic sampling of all the transversely sectioned bronchioles (less than 3:2 axial ratio in cross section) that satisfied these inclusion criteria was performed on each of four GMA-embedded sections per animal. Each airway was photographed (Zeiss Photomicroscope II) at a magnification adapted for it. As this sampling did not result in a sufficient number of statistical units per animal for comparative purposes. the sampled bronchioles were pooled and classified according to treatment. We measured the volume densities of the hronchiolar wall, lumen. and epithelial and muscular layers using the linear integration principle of Rosiwal (Weibel. 1979). From these measurements. the wall-to-lumen ratio and the epithelial mean thickness were calculated. We also determined the indentation index of the epithelium. so as to describe its degree of folding into the bronchiolar lumen. The analysis was done by projection of the micrographs onto a quadratic test grid ofdimensions 18.5 * 28 cm. comprising 17 lines of 28 cm each. The final magnification of the bronchiolar tissue varied between 340 and 540X. Again, the intercepts made by the tissue structures of interest were electronically measured (Zeiss MOP III). The indentation index was calculated as the number of intersections of the bronchiolar epithelial surface with the test lines system divided by the number of intersections of the basal lamina with the same system (adapted from Weibel, 1979). Stnfistical anu/j:srs. Analysis of the entire set of raw data obtained for the whole lung and the alveolar parenchyma in the exposed animals indicated different subtypes of response to chronic CD exposure. In an attempt to objectively classify these exposed animals according to their general pattern of response. we defined the normal limits of each of the stereological parameters measured for those two compartments as equal to the means + 2 SD. as measured in the 17 sham-exposed animals. This procedure revealed that the exposed animals could be classified, for the parameters considered, in one of two defined subtypes of responders: type I responders (within the normal value limits for most but not all parameters measured, II = 9) and type II responders (outside the normal value limits for all measured parameters. ~1= 8). The data were analyzed by a one-way analysis of vatiante for parallel groups, with multiple means comparison according to Duncan (Wirier. 197 I). The homogeneity of variances was verified through Bartlett and Cochran tests: when required. data were transformed by simple arithmetic functions to reach this homogeneity. The significance threshold was fixed a priori at p < 0.05,

RESULTS

The morphometric various histological

results obtained for the compartments of the

440

COULOMBE.

FTLION. TABLE

AND

CCjTB

I

Experimental

Parameters

Controls (N = 17)

I,‘,. par, lung VL. airw, lung VI, vast. lung

0.935 f 0.008b 0.023 _t 0.004 0.042 +- 0.005

Vlung (ml) Vpar (ml) brairw (ml) I’ vast (ml) Lung wet weight(g)

9.81 9.18 0.22 0.40 7.02

i 0.48 kO.48 + 0.04 f 0.05 f 0.20

group

Responders (N=9) 0.913 0.039 0.047 13.58 12.44 0.53 0.60 9.17

1

*0.011 + 0.006* * 0.008 k 1.06* * 1.01* + 0.10* k 0.09 -to.57*

’ V,-. volume density; V, absolute volume: par, parenchyma; airw, airways; vast, vascular ’ Data are expressed as Xk SE. ’ Water displacement volume ofcollapsed lungs and trachea at sacrifice. * Significantly different @ < 0.05) compared to control group. ** Significantly different @ < 0.05) compared to control and type I responders groups.

lung are shown in Table 1. The volume of the collapsed lung and its net weight were significantly increased in type I and type II responders. In type I responders, the proportion of the parenchyma and major vessels was normal, but the airways occupied a significantly increased fraction of the lung. Due to the increased lung volume, determination of the absolute volumes of these compartments showed values above normal in all cases, though not significantly so for the blood vessels. A difference in the magnitude of response between the two types of responders was seen at this first level of sampling of the lung tissue. As shown in Table 1, the increase in collapsed lung volume was much more in type II responders, for which there were significant alterations of the V, values of lung histological compartments: vessels and airways were increased while the parenchyma was decreased. Accordingly, the absolute volumes of these compartments were found to be altered compared to the controls. Histopathological examination indicated that the increased values measured for the airways in both subgroups and for the vessels in the type II responders were due to a thick-

Responders (N=8)

II

0.870 + 0.015** 0.060 k 0.009** 0.07 1 f 0.0 13* 16.54 14.35 1.01 1.18 9.70

+ 1.01** f 0X5* -+0.19** i- 0.24** i 0.4s*

compartment.

ening oftheir walls. This thickening appeared to be the result of a lymphocytic infiltrate and increased cellularity in their outer layers and not to a noticeable increase in stainable connective tissue fibers, as demonstrated by the Masson trichrome procedure. Alveolar Parenchyma Morphometric results of the alveolar parenchyma are shown in Table 2. In type I responders, slight increases in septal volume and in numerical density of intermediate pneumocytes (and accordingly in the index of alveolar epithelial repair) reached statistical significance when compared to the controls. Otherwise, all the parameters were very close to the control values. On a morphological basis, therefore, the alveolar parenchyma of most of the exposed animals could not be distinguished from the controls (cf. Figs. 1 and 2). As could be inferred from Fig. 3, the type II responders showed an increased reaction to chronic CD challenge, documented by many stereological parameters. The volume fraction, mean thickness, and absolute vol-

MORPHOMETRY

OF

COTTON TABLE

DUST-EXPOSED

441

LUNGS

2

EFFECTSOFCHRONICCOTTONDUST INHALATIONONMORPHOMETRICPARAMETERS OFTHE ALVEOLAR PARENCHYMA’ Experimental group Parameters

Controls (N= 17)

I .,.septum. par [‘septum (ml) S,.epit. par (mm-‘) S, epit, septum (mm-‘) S epit, (m’)

f 1.00 i 61.67 f 581.8 + 0.57 t

i Septum (gm) ‘PmFunct septum .I (slope) B (ordinate) h d macrophages hB pneumocytes II TV,intermediate pneumocytes .4lveolar epithelial repair index(%)

-t- 0.14 i 0.01 0.20 * 0.01 0.58 k 0.02 143.9 i- 14.3 192.5 -+ 9.6 16.1 * 1.9 8.05 f 0.89

0.104

0.008’ 0.07 1.17 22.6 0.03

Responders 1 (N=9) 0.103 f 0.005 1.39 t 0.14* 59.52 t 1.17 586.2 k 18.7 0.74 + 0.07

3.33 0.93

3.24 0.94

f f

0.10 0.00

0.21 + 0.01 0.57 141.0

184.2 28.3 13.08

-+ 0.02 f 2.2 k 7.4 f 3.7* f 1.28*

Responders II (Iv= 8) 0.325 + 0.073** 4.36 f 0.76** 48.36 +- 3.75** 212.1 + 45.0* 0.70 IL 0.08 9.24 0.61 0.30

t f +

1.63** 0.09** 0.02**

1.05 f 0.13** 148.2 f 31.3 393.2 28.6 7.77

f 5 1.8** + 5.8* f 1.94

’ 1, . volume density: S,-. surface density; S. absolute surface area: T. mean thickness; N*, numerical density: par. parenchyma; epit. epithelium. h Data are expressed as S f SE. * Signiiicantly different @ < 0.05) compared to control group. ** Significantly different @ < 0.05) compared to control and type I responders groups.

ume of the septum were increased severalfold as compared to controls. The slope (4) and the ordinate (B) of the frequency distribution of septal thickness also were significantly raised in these animals. Such a thickening of the septal wall contributed to significantly reducing the surface density of the alveolar epithelium, when expressed relative to the alveolar parenchyma and to the septum itself. This decrease could not be detected when this surface was expressed in absolute terms (m’), mainly because of the large increase in total lung volume already documented in Table 1. The numbers of intermediate and type II pneumocytes per unit surface area of the parenchyma were significantly raised, but the index of alveolar epithelial repair was not different from that of controls. The percentage of alveolar septum functional for gas exchange was substantially decreased, from 93% in controls to 6 1% in type II responders. The Masson trichrome staining procedure

confirmed that the thickening of alveolar septa in this exposed subgroup resulted from increased cellularity rather than from deposition of stainable connective tissue fibers. Bmnchiolar

Tissue

The morphological aspects of typical bronchioles from controls, type I responders, and type II responders are shown in Figs. 4,5, and 6, respectively. Slight quantitative changes characterized the walls of the sampled bronchioles in type I responders (cf. Table 3). The wall-to-lumen ratio was increased, though not significantly. In fact, only the volume density and the mean thickness of the epithelial layer were significantly greater than controls in this group. The reaction measured in type II responders was much more striking: the I’” value of the wall relative to the whole bronchiole was significantly increased: ac-

442

COULOMBE,

FILION.

AND

C8Tc

MORPHOMETRY

OF

COTTON

cordingly the wall-to-lumen ratio was increased as well. The epithelial changes were also more pronounced in type II responders: in addition to the V, and mean thickness values, the indentation index of the epithelium was increased, though the increase was not statistically significant. No change in the proportion of smooth muscle within the whole bronchiole was measured in either subgroup of exposed animals. On an absolute basis, the muscle layer was thickened, but not differently from the wall itself. Qualitative evaluation of the epithelium revealed hyperplastic and degenerative changes in the most severely affected animals, with some shifts in the epithelium’s pattern of organization from a typical pseudostratified arrangement to a stratified one. In the worst cases, such a trend resulted in a partial or even complete obliteration ofthe bronchiolar lumen. Bronchiolization of alveolar walls (presence of a cuboidal cell lining of airway epithelial origin on the septal surface of alveoli adjacent to bronchioles) was observed in seven of nine type I responders, and in all the type II responders. Many morphological features indicated that these cells were of bronchiolar epithelium origin rather than type II pneumocytes: a high nucleo-cytoplasmic ratio, the smaller size of vacuoles in the apical portion of the cell, the presence of cilia at the tip of some cells, the presence of neighboring dome-shaped cells reminiscent of Clara cells, and the occasional direct continuity with the bronchiolar epithelium itself (cf. Fig. 7). We also noticed an increased occurrence of hyperplastic nodules in the exposed animals, as compared to controls. These nodules,

DUST-EXPOSED

LUNGS

443

mostly cellular in composition, had a roughly spherical shape and were preferentially located nearby airways and vessels. Their cellular content was heterogeneous, and several mitotic figures could be observed (cf. Fig. 8). Their biological significance in relation to chronic CD exposure remains to be elucidated. DISCUSSION Very few reports on the lung morphology following chronic CD exposure exist in the literature. Moreover, most if not all of these studies were conducted on autopsy specimens from human byssinotic subjects, and focused primarily on the proximal airways. To our knowledge, no data exist on the morphology of the distal parenchyma in this occupational lung disease. Using a morphometric approach, Edwards et al. (1975, 1984) found increased volume proportions of mucus glands, cartilage, and muscle in the upper airways (lobar and segmental bronchi) of 43 byssinotic subjects. However, these authors could not relate such changes specifically to cotton dust exposure. In a more exhaustive study by Pratt et al. ( 1980). significant associations were found between CD exposure and both mucus gland hypertrophy and goblet cell metaplasia in proximal airways and bronchioles of cotton workers, but only when cigarette smokers were excluded. Very limited comparisons are therefore possible between those studies and the present one. further considering that our study was conducted in the guinea pig. Al-

FIG. 1. Light micrograph of alveolar parenchyma from a control guinea pig. Note the thinness of alveolar septa and the regularity of alveolar architecture. Glycol methacrylate (GMA) section, toluidine blue stain (X 160). AA, alveolar atrium: AS, alveolar space. FIG. 2. Light micrograph of alveolar parenchyma from a type I responder. No morphological change is apparent. GMA section, toluidine blue stain (X 160). AA, alveolar atrium: AS, alveolar space. FIG. 3. Light micrograph of alveolar parenchyma from a type II responder. The normal architecture is altered: the septa are thickened (arrowheads) due to hypercellularity, and the alveolar spaces are narrowed and even obliterated in some places (*). GMA section, toluidine blue stain (X 160). AA, alveolar atrium; AS. alveolar space.

444

COULOMBE.

FILION,

AND

COTI?

MORPHOMETRY

OF COTTON

DUST-EXPOSED

445

LUNGS

TABLE 3 EFFECTSOFCHRONIC COTTONDUSTINHALATIONONMORPHOMETRICPARAMETERSOF

THEBRONCHIOLES”

Experimental group Parameters I,‘,.wall, airways 1Ic.lumen, airways I’,-epit, airways I -i. must, airways Wall-to-lumen ratio T Epit (rm) I

Controls (N= 17) 0.597 0.401 0.200 0.134 2.30 25.47 1.77

+ 0.030h f 0.030 f 0.0 11 + 0.012 f 0.28 Tk 1.51 -+ 0.09

Responders I (N=9) 0.690 0.310 0.239 0.095 2.97 45.52 1.67

f 0.026 f 0.026 + 0.013* ?I 0.008 kO.43 zk 4.67* -+O.lO

Responders II (N= 8) 0.796 0.026 0.282 0.095 4.86 47.06 1.97

f 0.018** k 0.0 18** f 0.009** f 0.007 + 0.48 + 2.24* kO.13

’ I’,-. volumic density: 7. mean thickness: I, epithelium indentation index: epit, epithelium; must, smooth muscle layer. * Data are expressed as k + SE. * Significantly different (p < 0.05) compared to control group. ** Significantly different (p i 0.05) compared to control and type I responder groups.

though this small rodent may constitute a good model in inhalation studies, interspeties differences in airway geometry may influence the deposition of aerosolized particles into the various anatomical compartments of the lung. In particular, humans exhibit a relatively regular dichotomous pattern of airway branching, while most experimental animals possess an irregular dichotomous pattern termed monopodial (Schlesinger, 1985). In spite of these inherent limitations, Ellakani et al. (1984, 1985) and Alarie et al. (1985) have shown that the guinea pig provides a useful

animal model to study the effects of acute and chronic CD exposure, using pulmonary function tests. The morphometric approach used in this study permitted the identification, on a statistical basis, of two distinct patterns of lung response to chronic CD inhalation. This subclassification made use of the data from the whole lung and particularly the alveolar parenchyma. Data from airways were omitted, as the sampling procedure used did not cover the larger conductive airways of the guinea pig also involved in the pathogenesis of byssi-

FIG. 4. Light micrograph of a bronchiole from a control guinea pig. The wall to lumen ratio is low, and a normal pseudostratified epithelium is observed (inset). No sign of peripheral intlammatory infiltration is seen. GMA section, toluidine blue stain (X 160: inset X350). LUM. bronchiolar lumen: Ep, epithelium: m, smooth muscle. FIG. 5. Light micrograph of a bronchiole from a type I responder. Note the prominent lymphocytic infiltration (INF). An hyperplastic epithelium sometimes protrudes into the lumen (arrowheads), and at higher magnification shows degenerative changes (inset). The smooth muscle layer(m) is of normal thickness. GMA section, toluidine blue stain (X 160: inset X280). LUM. lumen; Ep, epithelium. FIG. 6. Light micrograph of a bronchiole from a type II responder. Inflammatory elements (INF) and edema surround the bronchiole. Increased epithelial thickness results in severe lumen (LUM) narrowing. The muscle layer is thickened in one location (arrowhead). Bronchiolization of alveoli (B,) can be seen adjacent to a terminal bronchiole (TB). Detail of the epithelium (inset) shows a stratified organization along with hyperplastic changes. GMA section. toluidine blue stain (X200: inset X350). Ep, epithelium; m. smooth muscle.

446

COULOMBE.

FILION.

AND

COTI?

MORPHOMETRY

OF COTTON

nosis (Edwards et al., 1984). On the basis of their physiological measurements, performed on the same guinea pigs either breathing normal air or challenged by 10% CO?, Alarie et al. (1985) were able to discern three patterns of lung response to this chronic challenge (see also Ellakani, 1985). These patterns consisted in lung restriction (3 animals), airways obstruction (4 animals), and a mixture of both ( 10 animals). There was no significant correlation between our morphological subclassification, the type I and type II responders, and the physiological patterns of response stated above. However, when stratified according to the physiological findings, the morphometric results obtained for the bronchioles correlated well with these (analysis not shown here). Further studies covering the entire tracheobronchial tree and conducted at the ultrastructural level are needed to adequately explore a possible correlation between morphometric findings and pulmonary function test results. Nevertheless, it can now be stated that both morphometric and physiological approaches were able to document differential susceptibility to chronic CD exposure. The calculation of a slope and an ordinate from the study of the frequency distribution of the septal thickness provides a new tool to evaluate the alveolar zone reaction to adverse stimuli. The information that can be inferred from these parameters is exemplified in the type II responders. As a first point, the fact that the mathematical transformations used in controls to linearize the frequency distribution still work with data from the exposed animals means that this distribution was still unimodal following treatment, and thus sug-

DUST-EXPOSED

LUNGS

447

gests on a statistical basis that the observed septal thickening was homogeneously distributed throughout the alveolar parenchyma following the CD toxic challenge. As a second point, a substantially increased ordinate, in the presence of an only slightly affected slope value (as is the case in type II responders). means that the entire population of alveolar septa in these animals was shifted toward higher thickness values, providing additional evidence that alteration in the distal lung was diffuse. Finally, the slightly (but significantly) raised slope measured in type II responders indicates that there has been a shift to the right in the septal thickness distribution as compared to the control group; this identifies an alteration of the normal pattern of distribution of this parameter in the lung. The total surface area of the alveolar epithelium, derived from our measurements in control guinea pigs, is in good agreement with previously reported values obtained through epoxy casting or stereological methods (Schreider and Hutchens, 1980). In type II responders, there was no significant change in surface area when expressed in absolute terms, whereas a substantial decrease appeared when it was expressed relative to either the parenchyma or the septum. Thus the interplay in anatomical reference spaces led to the detection of an effect in this subgroup of exposed animals, and serves to emphasize the usefulness of a multistage sampling approach for the present type of study. Stereological study of the oxygen conductance of the alveolo-capillary barrier has shown that a fivefold increase in its thickness can lead to a 50% decrease in conductance, thus significantly altering the gas exchange

FIG. 7. Bronchiolization of alveolar spaces in a type II responder. In many alveoli (ALV), bronchiolar epithelial cells (arrowheads) coexist with typical type II pneumocytes (II). These unusual alveolar cells share the same morphological characteristics and form a continuous lining layer with the epithelial cells of the bronchiole shown here (arrows). LUM. bronchiolar lumen. GMA section. toluidine blue stain (~700). FIG. 8. Light micrograph of a hyperplastic nodule from a type II responder. Located within the alveolar parenchyma, this nodule is composed of cells featuring nuclei of various shape, size, and density, contirming its heterogeneous cellular composition. At a higher magnification, mitotic figures can be seen (inset), GMA section. toluidine blue stain (X3 15; inset X720).

448

COULOMBE,

FILION,

process (Weibel, 1973). In the present study, it was assumed that changes in alveolo-capillary barrier thickness could be monitored through the measurement of the whole septal thickness, and also that “normal” oxygenation of the blood occurs when the morphometric estimate of the septal conductance for oxygen is above 50%. Although a normal total epithelial surface was maintained in the respiratory zone of type II responders, through tremendous lung enlargement, the percentage of alveolar septa still functional for gas exchange decreased therein to 6 1%. In one animal, this value was as low as 26% (data not shown). Thus, on a morphometric basis, the functional integrity of the respiratory zone of the lung was at least partially compromised in nearly 50% of the animals chronically exposed to CD (8 of 17). In type I responders, the relative number of type II pneumocytes was very similar to the control value, while that of intermediate pneumocytes, a transient form between type II and type I cells, was increased. Accordingly, the proportion of the total epithelial regenerative pool which was actively engaged in alveolar epithelium repair was raised by 62.5%. In type II responders, an exact opposite situation holds true: while the relative number of type II cells was approximately doubled, a normal percentage of these stem cells was engaged in a process of alveolar epithelial repair. These cell counts suggest that while the lungs of type I responders still display a potential for alveolar epithelium repair after chronic CD challenge, those from the type II responders have seemingly lost this capability, further considering that they are much more damaged. Whether this situation is a consequence of, or is partly responsible for, the marked pathological changes seen at the alveolar level remains to be elucidated. The results obtained from the analysis of bronchi from the 5th to 9th orders of the guinea pig respiratory tract indicate that these peripheral airways may be involved in byssinosis. Hyperplasia and degenerative changes of the bronchiolar epithelium occurred in

AND

COTi:

type I and particularly in type II responders, at the expense of the airway lumen. At this level of the respiratory tract, however, no overall change in the proportion of smooth muscle was observed in either subgroup, although a few individual bronchioles displayed thickening of the muscle layer. Moreover, bronchiolization was observed in most of the exposed guinea pigs. This feature was first characterized by Nettesheim and Szakal ( 1972) in adult mice, after long-term exposure to synthetic smog or CaCr04 dust. More recently, similar findings were reported by Fujinaka et al. ( 1985) following 1 year exposure of adult monkeys to 0.64 ppm of ozone. In CD-exposed animals, it occurred by proliferative expansion of bronchiolar epithelium through pores laterally connecting bronchioles and alveoli (cf. Fig. 7) just as described by Nettesheim and Szakal ( 1972). This may well constitute a mechanism of adaptation to chronic inhalation exposure to an adverse stimuli, whereby the very sensitive type I pneumocytes are replaced by more resistant bronchiolar epithelial cells. In conclusion, this study demonstrated that daily exposure to cotton dust for 1 year induced morphological changes in the parenchymal, airway, and vascular compartments of the guinea pig lung. Two subtypes of responders were discerned among the exposed animals, suggesting differential susceptibility in the development of byssinosis. In particular, the morphometric changes measured in distal airways and in alveolar septa of the most affected animals (8 of 17) both may contribute to a significant alteration of the primary function of the lung, the oxygenation of blood. The origin, in time, of these changes, and their development along the chronic process of alveolar injury, undoubtedly merits further investigation at the ultrastructural. biochemical. and functional levels. ACKNOWLEDGMENTS The authors are indebted to Dr. Ellakani. Dr. Karol, and Dr. Alarie ofthe University of Pittsburgh for submitting the lungs for analysis. Ms. G. Lassonde. Ms. M. Mor-

MORPHOMETRY

OF

COTTON

isset, Mr. J. Bernier, and Mr. D. Rodrigue are gratefully acknowledged for their skillful technical assistance. Special thanks are due to RhGne-Poulenc Pharma Inc. (MontrCal) for funding the purchase ofa microcomputer.

REFERENCES ADAMSON, 1. Y. R.. AND BOWDEN, D. H. (1974). The type II cell as a progenitor of alveolar epithelial regeneration: A cytodynamic study in mice after exposure to oxygen. Lab. Invest. 30, 35-42. ALARIE, Y., ELLAKANI. M., WEYEL, D.. AND KAROL, M. (1985). Respiratory parameters which characterize the response of guinea pigs to inhalation of cotton dust. In Proceedings, Ninth Cotton Dust Rex Cvnf: (P. J. Wakelyn and R. R. Jacobs, eds.). pp. 84-86. National Cotton Council and Cotton Inc.. Raleigh, NC. BERNIER. J.. FILION, P. R.. AND CGTE, M. G. (1984). Improved stability of a purified GMA preparation. SlUilf Technol. 59, 355. COULOMBE, P. A., FILION, P. R., AND C;)T& M. G. ( 1985). An histomorphometric assessment of BHT-induced pulmonary toxicity. J. Toxicol. Enviro?z. Health 15, 125-136. CRUZ-ORIVE, L. M.. AND WEIBEL. E. R. (1981). Sampling designs for stereology. J. Microsc. 122,235-257. EDWARDS. C.. CARLILE, A., AND ROOKE, G. (1984). The larger bronchi in byssinosis: A morphometric study. J.

C’lin. Pathol. 37, 20-Z. EIIWARDS. C., MACARTNEY. J.. ROOKE. G., AND WARD. F. ( 1975). The pathology of the lung in byssinotics. Tilora~- 30,6 I-1-623. EIUAKANI. M. A. (1985). .4n .4nimal .Wodel.for Bys.sin+ KS. Doctoral thesis. University of Pittsburgh. Pa. ELLAKANI. M. A., ALARIE, Y.. WEYEL, D., AND KAROL, M. (1985). Effects from 12 months inhalation of cotton dust. In Proceedings Ninrh Cotton Du.d Res Cw$ (P. J. Wakelyn and R. R. Jacobs, eds.), pp. 82-83. National Cotton Council and Cotton Inc., Raleigh. N.C. ELL~KANI. M. A.. ALARIE. Y. C.. WEYEL, D. A., MAZLJMDAR. S., AND KAROL, M. H. (1984). Pulmonary reactions to inhaled cotton dust: An animal mode1 for byssinosis. ToA-icol Appl. Phannucoi. 74, 267-284. EVANS, M. J.. CABRAL, L. J., STEPHENS, R. J.. AND FREEMAN. G. ( 1975). Transformation of alveolar type

DUST-EXPOSED

449

LUNGS

II cells to type I cells following

exposure

to NOL.

E.y

hIol. Pathol. 22. 142- 150. FUJINAKA, L. E.. HYDE, D. M.. PLOPPER. C. G., TYLER. D. L., DUNGWORTH, D. L.. AND LOLLINI, L. 0. (1985). Respiratory bronchiolitis following long-term ozone exposure in bonnet monkeys: A morphometric study. Exp. Lung Res. 8, 167- 190. HELANDER, K. G. (1983). Thickness variation within individual paraffin and glycol methacqlate sections. J. ,Wicrosc. 132 .- ?23-227. HOLMES, A. H. ( 192 I ). Petrqqruphic Mrrhod.y u& C’rriculations. Murby. London. LUNA, L. G. (1968). 12fanual qf Histological Stuining

.bfethod.yof the .4rmed Forces Institute of Pathology. McGraw-Hill. New York. NETTESHEIM. P.. AND SZAKAL, M. S. (1972). Morphogenesis of alveolar bronchiolization. Lab. Invexf. 26, 210-219. PRATT, P. C., VOLLMER, R. T., AND MILLER. J. A. ( 1980). Epidemiology of pulmonary lesions in nontextile and cotton textile workers: A retrospective autopsy analysis. Arch. Environ. Ifealrh 35. l33- 138. RYLANDER, R., HAGLIND, P., AND BUTCHER. B. T. ( 1983). Reactions during work shift among cotton mill workers. C/zest 84,403-407. SCHACHTER, N. E.. MAUNDER, L. R.. AND BECK, G. J. ( 1984). The pattern of lung function abnormalities in cotton textile workers. .Imr~r. Rev Respir. Dis. 129, 523-527. SCHLESINGER, R. B. (1985). Comparative deposition of inhaled aerosols in experimental animals and humans: A review. J. Toncol Envirox Health 15, 197-2 14. SCHREIDER, J. P.. AND HLJTC‘HENS, J. 0. (1980). Morphology of the guinea pig respiratory tract. .Inaf. Ret 196.313-321. SIMS.B. (1974). A simple method for preparing 1-2 Frn sections of large tissue blocks using glycol methacrylate. .I hlicrosc. 101,223-227. WEIBEL. E. R. (1973). The morphological basis of alveolo-capillary gas exchange. PhJsiol. Rev 53, 4 I9495. WEIBEL, E. R. (1979). Stereological kfeihods, Vol. 1. Practical klethods jiw Biolo,qical Akrphometrv. Academic Press, London. WINER. B. J. ( 197 I). Statisticat PrincipLes itf fzkperrmental Desi~Fn.McGraw-Hill. New York.