In situ microscopic analysis of asbestos and synthetic vitreous fibers retained in hamster lungs following inhalation.

Hamsters breathed, nose-only, for 13 weeks, 5 days/week, 6 hr/day, either man-made vitreous fiber (MMVF)10a, MMVF33, or long amosite asbestos at approximately 300 World Health Organization (WHO) fibers/cc or long amosite at 25 WHO fibers/cc. [World Health Organization fibers are longer than 5 microm and thicker than 3 microm, with aspect ratio >3.] After sacrifice, fiber burden was estimated (left lungs) by ashing and scanning electron microscopy (ashing/SEM) or (right middle lobes) by confocal laser scanning microscopy (CLSM) in situ. In situ CLSM also provided three-dimensional views of fibers retained, undisturbed, in lung tissue. Fibers of each type were lodged in alveoli and small airways, especially at airway bifurcations, and were seen fully or partly engulfed by alveolar macrophages. Amosite fibers penetrated into and through alveolar septa. Length densities of fibers in parenchyma (total length of fiber per unit volume of lung) were estimated stereologically from fiber transsections counted on two-dimensional optical sections and were 30.5, 25.3, 20.0, and 81.6 mm/mm3 for MMVF10a, MMVF33, and low- and high-dose amosite, respectively. Lengths of individual fibers were measured in three dimensions by tracking individual fibers through series of optical sections. Length distributions of amosite fibers aerosolized, but before inhalation versus after retention in the lung were similar, whether determined by ashing/SEM or in situ CLSM. In contrast, the fraction of short MMVF10a and MMVF33 fibers increased and the geometric mean fiber lengths of both MMVFs decreased by approximately 60% during retention. Most likely due to fiber deposition pattern and differences in sampling, fiber burdens [MMVF10a, MMVF33, and amosite (high dose; 269 WHO fibers/cc)] determined by ashing/SEM were 1.4, 1. 5, and 3.5 times greater, respectively, than those calculated from in situ CLSM data. In situ CLSM is able to provide detailed information about the anatomic sites of fiber retention and also fiber lengths and burdens in good agreement with ashing/SEM results.

The health effects of exposure to airborne asbestos and synthetic fibers have been extensively studied, as reviewed by DeVuyst et al. (1) and Hesterberg et al. (2). The respiratory system's response to fibers depends greatly on fiber composition, fiber length, the sites of deposition, the tissue compartments through which fibers move, and the rates at which they are cleared (3,4). Such health effects also depend on the amount of fiber deposition in the respiratory system and the resulting fiber burden. This has traditionally been determined by recovering fibers from lungs using thermal ashing techniques or chemical digestion by strong bases and oxidizing agents (5)(6)(7). Despite a considerable body of work, little is known about either the distribution of artificial fibers within the lungs after exposure or the local responses of lungs to fiber deposition. Indeed, it has been difficult to obtain such information because, although digestion methods permit determinations of numbers and dimensions of all fibers retained in the lung, neither fiber position within pulmonary structures nor any relationship to regional lung pathology can be recovered. Also, visualization methods available hitherto have required that the lung be finely sectioned physically to examine the fibers directly.
The mechanical properties of the fibers themselves make them difficult to approach without disturbing or breaking them; not only are they needlelike (inhaled fibers frequently have aspect ratios >5:1 and may be over 20 pm in length), they also may be both rigid and fragile. Thus, if samples of lung tissue are embedded in paraffin and physically sectioned in preparation for conventional transmission light microscopy, the fibers may be cut, disturbed, or torn out of the section during the cutting process, as fibers would typically not lie entirely within physical sections that may be 6-8 pm in thickness. Transmission electron microscopy (TEM) has been used to image fibers within fixed lung tissue (8,9), but here, especially, the process of obtaining physical sections for TEM (60-70 nm thick) can be expected to cut apart every fiber encountered and to introduce artifacts (6). Scanning electron microscopy (SEM) permits intact fibers to be studied, especially following freeze-etching or dissection along bronchi; however, SEM images show primarily the surfaces of fibers, cells, and tissue that are closest to the observer.
Confocal microscopy, however, makes it possible to peer tens or hundreds of microns below the surface of a translucent specimen and recover two-dimensional images from essentially any loc&ii)n within it, even revealing fibers within cells or tissues. Data volumes (three spatial dimensions plus intensity values) built up from such two-dimensional images obtained from increasing depths can be presented as stereoscopic (quasi-three-dimensional) pictures showing, for example, intact, undisturbed fibers penetrating alveolar walls.
The present study's purpose was to determine the numbers and locations of three types of fibers [two man-made vitreous fibers (MMVFs) and long amosite asbestos, a naturally occurring mineral fiber] that had been inhaled into hamster lungs. We utilized confocal microscopy to examine relatively large embedded lung tissue samples with minimal disruption of the fibers deposited there and compared the results to those obtained by SEM of ashing residue.

Materials and Methods
Experimental Design This article presents results obtained by examining material produced as part of a larger study. The parent study was an 83week chronic inhalation study with interim sacrifices; the first one, after 13 weeks of exposure, supplied the material for the present investigation. There were 125 hamsters in each of five exposure groups and 140 in an air control group. Some results from the parent study have been published (10). Experimental Material Fibers. Three types of fibers were studied: 1) MMVF 1 Oa, a glass fiber commonly used in building insulation; 2) MMVF33, a glass fiber used in the manufacture of highefficiency air-purification filter systems; and 3) long amosite, a form of naturally occurring mineral asbestos. World Health Organization (WHO) fibers have aspect ratio >3, length >5 pm, and diameter <3 pm (11).
Animals. Male Syrian golden hamsters (strain LakCrl:LVG) of specific-pathogenfree quality were housed and exposed to aerosolized fiber at Research Consulting Co. (Fiillinsdorf, Switzerland). They were housed individually in Makrolon type-1 cages (EHRET GmbH, Emmendingen, Germany) containing dust-free, softwood bedding (Lignocel, Schill AG, Muttenz, Switzerland) in a specific-pathogen-free environment. Pelleted food and water were provided ad libitum. This breed has been used in previous inhalation studies with natural and/or synthetic vitreous fibers (12).

Experimental Exposure
Animals were divided into six groups (five exposure groups and one air control group). Group 1 (MMVFlIa fibers) was composed of animals exposed to 339 WHO fibers/cc (29.6 mg/mm3); group 2 (MMVF33 fibers) was composed of animals exposed to 310 WHO fibers/cc (37.0 mg/mm3); group 3 (long amosite, low concentration) animals received 36 W-HO fibers/cc (0.8 mg/mm3); group 4 (long amosite, medium concentration) animals received 165 WHO fibers/cc (3.7 mg/mm3); group 5 (long amosite, high concentration) animals were exposed to 269 WHO fibers/cc (7.1 mg/mm3); and group 6 (air control) animals received no fibers, only air. For nose-only exposure to aerosols or filtered air, all animals were placed in restraint tubes for 6 hr/day, 5 days/week for 13 weeks; during the rest of the time, they breathed clean filtered room air. One day after the exposure period, hamsters were anesthetized with an intraperitoneal injection of sodium pentobarbital (300 mg/kg body weight) and sacrificed by exsanguination, whereupon their lungs were excised.

Specimen Preparation for in Situ Microscopic Examination
Fixation and embedment. The middle lobe of the right lung of each hamster was processed for microscopic examination of fibers in situ. The right lung was detached between lung and hilum, fixed by instilling 4% buffered paraformaldehyde into the airways at 30 cm water pressure, then shipped from Switzerland to Boston. The lung samples remained in the fixative for up to 10 days. Each right middle lobe was cut sagitally into five pieces 5 mm thick, dehydrated in graded ethanolic series to absolute ethanol, stained nonspecifically with the fluorescent dye lucifer yellow CH [0.001% w/v in absolute ethanol; (13)], and embedded in Spurr's epoxy (Polysciences, Warrington, PA). The results of Law et al. (14) suggest that at most 16% of the silica content of the most susceptible MMVFs would be dissolved in this time, and that the loss could be less than 10% from all fiber types.
Sample manipulation. Epoxy-embedded samples were attached to a metal adapter plate and mounted on a highspeed milling apparatus, where the top of the block was milled to a glassy smoothness (15). The sample on its adapter plate was then attached to a stage with digital readouts on the confocal laser scanning microscope/microscopy (CLSM). The CLSM can only retrieve optical sections to a certain depth, the depth limit depending on objective working length and the severity of light scattering and absorption by the sample. Material below this depth limit could be imaged by returning the sample block to the milling apparatus to remove some of the already imaged overlying tissue. Upon its return to the CLSM, the block could be positioned in three dimensions to within 1 pm of its previous position.

Specimen Preparation for Ash Analysis
The left lung (detached between lung and hilum) was frozen at autopsy and shipped frozen from Switzerland to the Johns Manville International, Inc., analysis laboratories (Littleton, CO). To recover inhaled fibers from the lung for these evaluations, tissue from the left lung was subjected to a low-temperature ashing process as previously described (7,10,16); the ash residue was suspended in deionized filtered water, sonicated, and filtered onto a polycarbonate membrane filter (Nucleopore; Bio-Rad Laboratories, Hercules, CA; 0.2 pm pore size), which was mounted on a specimen planchet, coated with gold, and examined by SEM. Fibers were counted and measured according to methods described previously (10 The CLSM was configured with two channels, each with a separate photomultiplier detector. In channel 1, used for fluiorescence (tissue) detection, light (wavzelengths >510 nm) emitted from dyestained tissue was directed to detector 1. In channel 2, used for reflected light (fiber) detection, illuminating light reflected from fibers (wavelengths <510 nm) passed through a 510-nm secondary beam splitter and was directed to detector 2.
Each field-of-view (FOV), therefore, was represented by two digital imagesone from each channel and such pairs of images could be superimposed to show fiber positions in relation to lung tissue. Optical sections lay in the microscope's (x,y) focal plane. Fiber samples: imaging by CLSM. Samples of dry stock MMVF 1 Oa, MMVF33, and long amosite asbestos fibers were obtained from the NAIMA (North American Insulation Manufacturers Association, Alexandria, VA) fiber repository and were suspended (1 mg/ml) in saline and allowed to settle in microwell chambers for CLSM. Micrographs of fibers were obtained with the CLSM configured as described previously, using the 20x and 60x objectives.
Lung tissue and fibers retained in lung: imaging by CLSM. The CLSM was used to look through the smooth surfaces of the epoxy-embedded tissue samples to as much as 150 pm below the surface without physically sectioning the samples.
Stacks of aligned optical sections were acquired between two depths sufficiently separated to completely contain essentially all fibers detected within them. This technique permitted us to determine the length densities, length distributions, and anatonmic placement of retained long amosite and MMVF fibers in lungs. Three-dimensional projections were generated from stacks of optical sections using Voxel View software (Vital Images, Inc., Fairfield IA) (17).
CLSM imagesfor stereologic length densities. Length density refers to the length of a linear structure present within a unit volume of some containing phase; here, it refers to the total length of retained fiber per unit volume of lung and has units such as millimeters per cubic millimeter.
Micrographs were obtained with the CLSM configured as described previously, using the 20x objective. Each digital micrograpl was a two-dimensional, 1,024 x 1 024 arra-Volume 107, Number 5, May 1999 * Environmental Health PerspectivrV 520.0 pm on a side of gray-scale pixels (intensity values from 0 to 255), 0.5078 pm on a side; thus, pixel area was 0.2579 pm2 and image area 270,399 pm2. FOVs were acquired at random locations over the tissue exposed at the prepared block face.
CLSM images for three-dimensional lengths of individual fibers. Each image, acquired with the 60x objective, was a 512 x 512 array 87 pm on a side of gray-scale pixels, 0.34 pm on a side. The images were obtained with a vertical spacing of 0.7 pm. Stacks encompassing >75 pm along the zaxis were obtained from locations where fibers were relatively plentiful.
Image segmentation to emphasizefibers in the presence of tissue. Prior to analysis, each pixel in the field of view was assigned to one of three categories: tissue, fiber, or other. This was done by setting pixels representing air and noise in the channel 1 (tissue) image to 0, leaving tissue pixels with values >0. Fiber pixels in channel 1 were 0-valued. The channel 2 (fiber) image was similarly thresholded, leaving >0 values in pixels representing fiber; because of the slight reflectivity of the tissue and weak fluorescent emissions occurring at short wavelengths, some tissue appeared faintly in channel 2. Pixels representing fibers could be distinguished as those that were 0 in channel 1 but of at least moderate intensity in channel 2; side-by-side comparisons of the channel 1 and 2 images made it easy to identify fiber profiles in the channel 2 image. Length Densities of Fibers within the Lung, by in Situ CLSM Fiber length densities (total length of retained fiber per unit volume of lung) were measured in four, five, five, and five animals from the MMVFl0a, MMVF33, low-dose amosite, and high-dose amosite treatment groups, respectively (13-week exposure; no recovery period). Images were taken predominantly of the two tissue samples with large cut surfaces on both sides and were collected at random (regular grid with randomly positioned origin); exception to this random procedure was made in that FOVs with visible pleura or extrapulmonary space were not recorded. Fiber transsections visible on each image were counted manually. IA'   figure). At least 500 pm of epoxy with embedded tissue was milled away from the surface (see text) prior to examination to ensure that fibers seen here could not have been disturbed by the act of cutting samples from the fixed lung before embedment. (A) Man-made vitreous fiber (MMVF)10a: 31 fiber transsections seen in alveolar parenchyma; 21 are >5 pm from nearest septal profile. (B) MMVF33: 31 transsections of fibers (arrowheads) are seen within alveolar parenchyma; of these, 21 are more than approximately 5 pm from the nearest alveolar septal profile. Lung pleura is visible. (C) Long amosite: seven fiber transsections (arrowheads) and two alveolar macrophages (arrows) can be seen; one of the macrophages contains at least three fibers, the other, at least one. Bars represent 50 pm.
We assume that fibers retained in lung are randomly oriented with respect to the optical plane of section. The total length of linear features of interest (fibers) contained within a phase (lung) may be estimated (18)

Fiber Dimensions
Fiber dimensions before aerosolization, as determined by SEM. Lengths and diameters (analyzed using SEM and reported as mean, standard deviation, minimum, maximum, and median) of fibers to be aerosolized were obtained from NAIMA.
Dimensions of retainedfibers, as determined by ashing/SEM. Numbers and dimensilons of fibers after aerosolization, inhalation, retentioni, anid ashing of the left lung were determined by analysis of SEM micrographs of fiber samples collected on Nucleopore polycarbonate filters (0.2 pm pore size) (19). Fiber lengths as determined by CLSM in situ. The lengths of fibers retained in animals' right lungs were determined from fibers contained in stacks of optical sections. The length of a fiber was estimated as the sumn of Pythagorean distances between transsections on successive optical sections, whose locations were marked in three dimensions (pixel located at coordinates x,y on section at depth z).

Fiber Burden
Fiber burden calculatedfrom CLSM data. The number of fibers in parenchyma per lung was estimated by combining estimates of fiber length density and geometric mean fiber length with an assumed lung volume. (Lung volume was not measured directly.) where FB is the fiber burden, number of fibers per left and right lung pair; LV is the length density of fibers in lung, from twodimensional CLSM data and stereology; V is the volume of lung, assumed to be 4.0 ml; be is estimated geometric mcan length of fibers, from three-dimensional CI SM data.

Characteristics of Fibers before Aerosolization
Nonaerosolized samples of the three fiber types were examined by reflected light confocal microscopy. In Figure IA (MMVFlOa), the diameters along any given fiber appeared constant, but the fibers themselves occasionally were curved. Fibers occurred with a range of diameters, however. In Figure 1 B (MMVF33), individual fibers appeared straight, often aggregated side to side and jumbled together. In Figure IC (amosite), fibers appeared as straight rods of constant diameter, rarely clustered. All fiber types displayed brightnesses that varied depending on fiber orientation relative to the imaging plane, and they revealed interference patterns characteristic of fibers illuminated by polarized light. Practically the full lengths of the fibers were seen in these images, despite the optical sectioning property of confocal microscopy, because the fibers lay on a flat surface parallel to the optical sectioning plane. Because all fibers seen were in focus simultaneously, they must have been coplanar, and there was no indication of out-of-plane unobserved fibers. Hence, it is most likely that all fibers were clearly imaged and observed.

Placement of Retained Fibers
Samples of hamster lung following 13-week inhalation exposure to the three fiber types were examined using combined fluorescence and reflected light confocal microscopy. Fibers of each type reached the alveolar parenchyma and in some instances were taken up by alveolar macrophages (Fig. 2). Both MMVF1Oa ( Fig. 2A) and MMVF33 (Fig. 2B) appeared throughout alveolar parenchyma, even occurring in subpleural regions. Fiber transsections often were seen some tens of microns away from the nearest septal profile, meaning that the fiber did not lie against an alveolar septum for its entire length but crossed an alveolar airspace. With Volume 07 Number 5, May 999 * Environmental Health Perspectives 1-Af,b high-dose amosite (Fig. 2C), some fibers were found within alveolar macrophages (arrow, Fig. 2C).
Fibers of each type had reached alveolar parenchyma. Fiber placement within parenchyma was further established by examining stereoscopic views of stacked, aligned optical sections. For MMVF1Oa (Fig. 3A, 3B), fibers were present both free in the airspace and partially or wholly ingested by inflammatory cells. For MMVF33 (Fig.  4A, 4B), as with MMVFlOa, some fibers were engulfed by inflammatory cells within alveolar air spaces (Fig. 4A, hollow arrow). A long fiber lay within the interstitium of an alveolar septum (Fig. 4A, arrow). The accompanying stereo projection (Fig. 4B) shows fiber position and orientation in alveoli. In amosite (Fig. 5A, SB), a fiber (Fig. 5A, arrow) appears within the interstitial space adjacent to an airway. The stereo projection (Fig. 5B) shows this fiber to be entering the interstitium at an airway bifurcation (Fig.  5B, arrow). Numerous inflammatory cells containing fibers are present in the airway.
An alveolar macrophage can be seen attempting to engulf a long amosite fiber in Figure 6, which was acquired at higher magnification than Figures 3, 4, and 5. Most of the amosite fiber's length is either embedded in septal tissue or surrounded by the macrophage's cell body or its slender tapering processes. Table 1 summarizes lengths of MMVF lOa, MMVF33, and long amosite asbestos fibers as measured in air and lung samples 1) present in the aerosol presented to the animal's nose; 2) after 13 weeks of exposure, by the ashing/SEM technique; and 3) after 13 weeks of exposure by in situ CLSM.

Length Distributions of Fibers as Aerosolized and after up to Thirteen Weeks of Retention
The order of aerosolized fiber lengths from longest to shortest (for both geometric mean lengths and median lengths) was MMVFlOa, MMVF33, amosite; for retained fibers, the order was differentamosite, MMVFlOa, MMVF33. Both MMVF types were shortened by approximately 60% (geometric mean). The geometric mean for amosite increased by 44%, whereas the arithmetic mean and median decreased by 9 and 6%, respectively. All fiber length distributions are skewed toward longer lengths (see below). Figure 7 compares fiber length cumulative frequency distributions (FLCFDs) 1) for fibers in the aerosol, determined using SEM; 2) for retained fibers, using ashing/SEM; and 3) for retained fibers, using in situ CLSM. With MMVF1Oa (Fig. 7A), both analyses of retained fibers yielded similar FLCFDs and both found fewer long fibers than did aerosol analysis; in particular, 37% of fibers in the aerosol were longer than 20 pm while barely 5% of retained fibers were. With MMVF33 (Fig. 7BY,-ashing/SEM and in situ CLSM did not yield--similar FLCFDs for retained fibers; the former showed a smaller proportion of short (<5 pm) fibers than did the latter. Both, however, showed fewer long fibers (>15 pm) than did aerosol analysis. With amosite asbestos (Fig. 7C), ashing/SEM analysis of retained fibers showed more short (<5 pm) and fewer long (>20 pm) fibers than did in situ CLSM. Neither analysis of retained fibers, however, showed as many long fibers as did aerosol analysis. Figure 8 shows the complete FLCFD obtained by in situ CLSM for retained fibers of each type. The fraction of total fiber number lying between any two lengths may be read directly from Figure 8. For example, the proportion of fibers categorizable as respirable according to WHO standards (identical with the category of all fibers longer than 5.0 pm and listed in Table 2) may be obtained as the fraction of the ordinate above the crossing point of the FLCFD and the vertical dotted line at 5.0 pm. Similarly, these curves demonstrate that the proportion of long (>20 pm) fibers was less than 5% for MMVFlOa, nearly 15% for MMVF33, and 25% for amosite. Table 3 gives the number of FOVs examined, the number of fiber transsections found, and the calculated length density (Eq. 1) for each fiber treatment group. MMVFlOa and MMVF33 apparently . It is seen to be located at an airway bifurcation. The numerous inflammatory cells present in the airway (e.g., asterisk) contain many short and long fibers. Note the presence of long straight fiber in both airway and alveolar airspace (hollow arrow); by examining the individual sections and viewing the reconstruction from different angles, this fiber can be seen to penetrate the tissue bounding the alveolar duct.

Length Densities of Fibers in Lung by in Situ CLSM
behaved similarly during retention, according to their distributions (given previously) and Table 3 demonstrates that length densities of these fibers were also comparable.

Fiber Burden
Fiber burdens (Eq. 2) of MMVFlOa, MMVF33, and amosite (high dose) at the end of the 13-week exposure were estimated by two different techniques (ashing/SEM and in situ CLSM). Results are presented in Table 2, which lists the fraction of total fiber number in each 5-pm-wide length class up to 20 pm and the >20 pm length class; these numbers are the same as plotted in Figure 6 and were used to calculate the corresponding fiber burdens. The ratios between the fiber burdens for the two techniques are also provided for each fiber type. Discrepancies between the burdens are greatest for amosite. The burdens for amosite were two-to ninefold greater with SEM than with CLSM, with the short fiber category (<5 pm long) showing the greatest disparity; in millions of fibers/lung, the burdens were 17 by SEM as compared to <2 by CLSM ( Table 2). The two methods yielded fairly similar burdens for the two MMVFs, with SEM burden no more than 2.4-fold greater than CLSM. The distribution of the fiber burden by length class was similar for both MMVFs, except that MMVF33 had a greater number of long (>20 pm) fibers than did MMVFl Oa.

Locations of Fibers Retained in Lung, by in Situ CLSM
We observed the placement of MMVF and asbestos fibers in situ in lung tissue after 13 weeks of exposure. Fibers of all three types were deposited in small airways and alveolar septa, frequently being seen at airway bifurcations; this agrees with observations of Brody et al. (8) and Brody and Roe (20). Long amosite asbestos fibers frequently penetrated into alveolar septa as far as the interstitium and even completely through septa, in accord with the demonstrated ability of fibers to penetrate lung tissue (21)(22)(23). MMVF fibers were not observed to penetrate septa as frequently as did amosite fibers, however. Fibers of all types were found within alveolar macrophages.

Effects of Length-biased Filtering during Respiration and Fiber Fragmentation and Degradation during Retention
As in previous rodent inhalation studies (24), the present study demonstrated that both arithmetic and geometric mean lengths of MMVF fibers in aerosol were typically longer than those of retained fibers. This difference could be the result of a lung filtering process in which longer fibers are prevented from being inhaled into the distal alveolar parenchymal portions of the lung, or it could be a result of fiber degradation within the lung, or both.
In the case of amosite (Fig. 7C), the aerosol FLCFD and retained FLCFD obtained by in situ CLSM were similar, suggesting that amosite fibers of all lengths were equally likely to reach parenchymathat is, that no significant filtering process affected amosite fibers. Because the FLCFD of retained amosite fibers was obtained after 13 weeks of exposure-adequate time for fiber fragmentation and degradation to have occurred in the lungit seems that amosite fibers also resisted breakage and degradation and remained intact.
MMVFlOa and MMVF33, on the other hand, showed strong reductions in arithmetic and geometric mean lengths (Table 1) and a strong increase (Fig. 7A, 7B) in the proportion of short fibers over the same time period. The 13-week time point cannot shed much light on which process (filtration or degradation) is more important in causing these changes because the possible filtration effects could be masked by effects of degradation. However, as part of the chronic study, fiber lengths were determined after only 6 hr of retention (10,19), and the proportion of long fibers by fiber type was as follows: MMVF1Oa: aerosol>6 hr>13 weeks, suggesting that both lung filtration and fiber degradation were contributing factors; MMVF33: aerosol>6 hr = 13 weeks, suggesting filtration but not degradation; amosite: aerosol = 6 hr = 13 weeks, suggesting neither filtration nor degradation. The absence of a filtration effect on amosite fibers might be due to the narrower fiber diameters of that mineral as compared to the MMVFs, which could give long amosite fibers a smaller, hence more respirable, aerodynamic diameter than long MMVF fibers, which might have an aerodynamic diameter large enough to make them less respirable.

Comparison ofTwo Methods for Measuring Lung Fiber Burden
The literature suggests that destroying the lung to assay retained fibers may change their lengths; Warheit et al. (6) showed that glass fiber dimensions become modified during simulated tissue digestion, possibly leading to an overestimation of the numbers of fibers/lung (e.g., as a result of With only two exceptions (MMVF 1 Oa and MMVF33 fibers <5 pm long), fiber burdens estimated by ashing/SEM were greater than those estimated by in sitl CLSM by factors of 1.4-9.1 ( Table 2). The largest difference in counts (ninefold) between the two methods was for amosite fibers <5 pm in length (Table 2). At least two parameters. fiber diameter and fiber distribution throughout the lung, could account for the reduction in the numbers of fibers detected by the in situ CLSM method relative to the ashing/SEM method. The images in this study were collected using the CLSM in the reflected light mode and had pixels approximately 0.51 pim on a side.
Fibers with diameters <0.5 pnm may not have reflected enough light to raise the intensity of the image's correspondinig pixel(s) above noise level. This coulld explain the especially large (twoto ninefold) discrepancies between SEM and ClSM counts for amosite, which had smaller mean diameters than the two MMVFs. Additionally, in situ CLSM would detect fewer fibers than ashing/SEM if fiber deposition was not homogeneous through the lung. For example, during sparse random imaging of a lung sample by CLSM, it would be relatively rare to encounter fibers cluimped in a fesv dense local concentrations; however, with ashing/SEM analysis, all fibers---whethei cltumped or isolated would be equally likely to be counted. Finally, fibers approximately 5 pm long can be phagocytized by alveolar macrophages and cleared to pillmonary lymph nodes, where they may remain for an indefinite period (25). Fibers within lymph nodes would have beeni detected by ashing/SEM, but in the present study no lymph nodes were examined by in situ CLSM. This difference between the two methodologies may adequately account foimany of the observed discrepancies. transverse fragmentation). In the present study, the in situ CLSM method provides data on fiber length free from any artifacts due to lung destruction. With MMVF1Oa, in situ CLSM produced a postretention FLCFD that overlaid that from ashing/ SEM analysis ( Figure 6A), suggesting that ashing did not fragment retained MMVF1Oa fibers significantly. Results from MMVF33 run counter to the breakage-during-ashing hypothesis; the FLCFD from in situ CLSM shows a greater proportion of short fibers than that from ashing/SEM. We cannot yet account for this discrepancy. The amosite results, however, were consistent with breakage during ashing; the FLCFD by in situ CLSM showed a greater proportion of long fibers than the FLCFD by ashing/SEM, and the numbers of actual fibers/lung determined by SEM were approximately 250% greater than those determined by in situ CLSM. However, other differences in the two analytical methodologies may offer a better explanation for the discrepancies in amosite counts.
There is room for technical improvements, notably in the increase and calibration of detection sensitivity. Effort also must be made to ensure that fibers are not unduly disturbed during tissue preparation and embedding and that tissue configuration remains reasonably unchanged. In addition, in situ analyses are time-and labor-intensive, although advances in speed of image acquisition and subsequent computation continue to ameliorate this situation. To its credit, in situ CLSM analysis brings several advantages to the study of fiber toxicity models. Because retained fibers may be visualized largely undisturbed, their locations relative to cells and surrounding tissue cani be stLidied anid Volume 107, Number 5, May 1999 * Environmental Health Perspectives  understood. Some progress toward this understanding can be made using the ashing/SEM technique; lungs may be lavaged, and the lavage fluid may be centrifuged so as to pellet the cells (primarily alveolar macrophages) but not the free extracellular fibers, which then remain in the supernatant. Ashing/SEM may be used to examine the fibers in the lavaged lung and the lavage pellet, producing estimates of the fibers that penetrated lung tissue and those which macrophages had engulfed or attempted to engulf. However, this technique is vulnerable to the uncertain nature of lavaging; it is not clear, for instance, whether all fibers in alveolar air spaces would be swept out into the lavage fluid or whether they could remain trapped.
Additionally, ashing/SEM still does not provide information on where and how fibers and tissue interact or on whether macrophages ingest fibers successfully. In situ CLSM analysis, as shown here, does provide information on these points.
Three-dimensional reconstructions of fibers and tissue can be particularly informative about the modes by which fibers mechanically damage tissue and the mechanisms the organ uses to respond to the insult; in the present study, for example, fibers penetrated septa and airway walls and were partially or entirely engulfed by phagocytes. It would seem possible, also, to use three-dimensional microscopy to study whether fibers and their fragments migrate within tissue or are carried within macrophages, possibly to be cleared from the lung.