The alveolar epithelium is a mosaic of type I cells that build the ultrathin air-blood barrier and type II cells that synthesize and secrete surfactant. Surfactant is composed of lipids (∼ 90%) and proteins (∼ 10%) and maintains alveolar integrity by reducing the surface tension at the air-liquid interface. The surfactant proteins A (SP-A) and D (SP-D) are members of the collectin protein family, and they have comparable potential to interact with cells and microorganisms (Crouch and Wright, 2001; Hawgood and Poulain, 2001). The lung collectins, SP-A and SP-D, are part of the pulmonary innate immune system (Crouch and Wright, 2001; McCormack and Whitsett, 2002; Wu et al., 2003; Wright, 2004) and influence inflammatory processes (Chabot et al., 2003; Gardai et al., 2003). Within the alveolar lumen, SP-A is associated with tubular myelin, whereas SP-D exists mainly free in the hypophase (Wright, 1997; McCormack and Whitsett, 2002). The distribution in different intra-alveolar microdomains may possibly lead to distinct functions.
Gene-targeted mice lacking SP-A, SP-D, or both are vital, fertile, and free of infections when unchallenged (Korfhagen et al., 1996; Botas et al., 1998; Hawgood et al., 2002). SP-D single deficient mice show altered lung pathology consistent with emphysema, abnormalities of alveolar macrophages, as well as of type II cells and an increased alveolar surfactant pool size (Botas et al., 1998; Ikegami et al., 2000a; Wert et al., 2000). SP-A single deficient mice lack tubular myelin but have no obvious changes in surfactant metabolism under resting conditions (Korfhagen et al., 1996; Ikegami et al., 1997).
The development and first characterization of mice deficient in both SP-A and SP-D revealed progressive intra-alveolar accumulation of phospholipids and proteins. The initial description also demonstrated emphysema-like pathology (Hawgood et al., 2002). The present study extends these observations and provides a detailed design-based stereological analysis at the level of light and electron microscopy to characterize the phenotype of lung collectin-deficient mice. Recently established methods for the direct and unbiased estimation of the number and size of alveoli (Hyde et al., 2004; Ochs et al., 2004a) and of the number and size of type II cells as well as their surfactant-storing lamellar bodies (Ochs et al., 2004b) were chosen to provide new data on lung parenchymal architecture, on the morphology of alveolar type II cells and their lamellar bodies, as well as of alveolar macrophages in mice double deficient in SP-A and SP-D.
MATERIALS AND METHODS
The generation of the mice and the initial phenotype characterization were previously described in detail (Hawgood et al., 2002). Mice deficient in SP-A and SP-D were generated by sequential targeting of the closely linked genes in F1 B6/129Sv ES cells. All experiments described here were conducted on littermate mice of B6/129Sv:CD-1 mixed genetic background. Mice were housed in conventional cages and bedded with Paper Chip brand laboratory animal bedding manufactured (softer texture) by Sheppard Speciality Papers. Water was acidified, diet was Lab Diet 5053 Irradiated PicoLab Rodent Diet 20. The following pathogens were excluded by the Laboratory Animal Resource Center sentinel program: mouse hepatitis virus, Sendai virus, pneumonia virus of mice, mouse parvovirus, minute virus of mice, Theiler's murine encephalitis virus, and mycoplasma pulmonis. Study mice were wild-type for SP-A and SP-D (WT) or null for both genes (A−D−). Mice were genotyped by PCR using primers specific for the SP-A and SP-D alleles. Experiments were conducted on mice of 12 weeks of age, n = 5 per genotype. The reason for choosing five animals per group in a stereological study is that if a parameter is found to change in one direction in all five cases, then the probability that this is due to chance is P = (1/2)5 < 0.05, thus making the experiment conclusive (Cruz-Orive and Weibel, 1990). All experimental protocols were approved by the Institutional Animal Care and Use Committee.
Fixation, Sampling, Processing
Mice were euthanized by intraperitoneal injection of pentobarbital (200 mg/kg) prior to bilateral thoracotomy. After instillation fixation of 2% glutaraldehyde and 1% paraformaldehyde in 0.1 M phosphate buffer via the trachea at a pressure of 20 cm H2O, the total lung volume was estimated by fluid displacement according to Archimedes' principle (Scherle, 1970) followed by systematic, uniformly random tissue sampling and postfixation as earlier reported (Fehrenbach and Ochs, 1998). For light microscopy, lung slices were osmicated, bloc-stained, dehydrated, embedded in glycol methacrylate (Technovit 7100; Heraeus Kulzer, Wehrheim, Germany), and cut into 1 μm sections stained either with methylene blue or with orcein. The remaining lung slices were cut into blocks, osmicated, bloc-stained, dehydrated, embedded in LX 112 (Ladd Research Industries, Burlington, VT), and cut into 1 μm semithin sections stained with toluidine blue for light microscopy and 100 nm ultrathin sections stained with uranyl acetate and lead citrate for electron microscopy.
The stereological methods applied are free of assumptions on the shape, size, orientation, or distribution of the structures investigated and therefore fulfill the criteria for design-based or unbiased stereology. Point and intersection counting (Weibel, 1979) estimated volume and surface densities, respectively. The physical disector (Sterio, 1984) was used to estimate numerical densities. Densities were multiplied by their respective reference volumes to obtain total values. The mean particle size was estimated by dividing the total volume of particles by their number. Furthermore, local stereological estimators, the planar rotator (Vedel Jensen and Gundersen, 1993) and the point sampled intercepts method (Gundersen and Jensen, 1985), were used to estimate the number-weighted mean volume and the volume-weighted mean volume of particles, respectively. The volume-weighted mean volume results from the number-weighted mean volume amplified by the coefficient of variation of the individual particle volumes and therefore contains information on both mean particle size and variability of size. For surface and size estimators, global isotropy was assumed. The contributions of the interindividual biological and the intraindividual methodological variations to the total observed coefficient of variation were analyzed to check if the stereological estimates were sufficiently precise.
For all outcome measures, four pairs of sections per lung were used. The light microscopic analysis was performed on an Axioskop light microscope (Zeiss, Oberkochen, Germany) connected with the computer-assisted stereological system CAST 2.0 (Olympus, Ballerup, Denmark). On pairs of orcein sections [distance: 3 μm; primary magnification (PM): 5×], the alveolar number was estimated as recently described for different mammalian lungs (Hyde et al., 2004; Ochs et al., 2004a, 2004b). From methylene blue sections (PM: 5×), the volume density of parenchyma in the lung and the volume density of alveoli in the parenchyma were estimated and multiplied by the lung volume, providing the total volume of alveoli. The mean size of an individual alveolus was determined by dividing the total volume of alveoli by their number. From pairs of toluidine blue sections (distance: 3 μm; PM: 63×; oil immersion), the number and the number-weighted mean volume of type II cells as well as of alveolar macrophages were estimated.
The ultrastructural stereological analysis (PM: 3,000–4,400×) was performed using a LEO EM 900 transmission electron microscope (Zeiss) combined with an image analysis system (AnalySIS; Soft Imaging Systems, Münster, Germany). Following systematic uniformly random sampling on pairs of ultrathin sections (distance: 100 nm), micrographs of corresponding type II cell profiles were taken. The volume densities of nuclei, mitochondria, and lamellar bodies as well as the apical, basolateral, and total cell surface density were estimated. All densities were multiplied by the mean type II cell volume, resulting in the total volume of nuclei, mitochondria, lamellar bodies, and the apical, basolateral, and total surface area per cell. Furthermore, the number of surfactant-storing lamellar bodies, their number-weighted mean volume, as well as their volume-weighted mean volume were estimated in type II cells. As a measure of the intracellular surfactant pool, as defined by morphological criteria, the total volume of lamellar bodies per lung was calculated as the product of the number of type II cells and the volume of lamellar bodies per cell.
Data are given as mean (CV) of n = 5 mice per genotype, with CV = SD/mean. The nonparametric two-sided Mann-Whitney U-test was used to analyze the stereological data statistically. The test was performed using the Statistica 6.0 software (StatSoftInc, Hamburg, Germany). A value of P < 0.05 was considered to be significant.
Stereological data are summarized in Table 1, and light and electron micrographs illustrating the quantitative findings are shown in Figure 1. At the level of light microscopy, the morphological analysis revealed a decreased number and an increased size of alveoli in A−D− mice compared to WT (Fig. 1a and b). Furthermore, the numbers and sizes of type II cells as well as alveolar macrophages were increased in A−D− mice (Fig. 1c and d). The ultrastructural analysis showed no differences in the volumes of type II cell mitochondria or nuclei between the two groups. The total surface area of type II cells in A−D− mice was enlarged with a significant increase in apical as well as basolateral surface area. In both groups examined, the two surface proportions contributed equally to the whole surface area (data not shown). The total lamellar body volume per cell was significantly increased in the type II cells of A−D− mice. There were more lamellar bodies per type II cell in the A−D− group. The mean volume of a single lamellar body was not statistically different. The significant increase in the volume-weighted mean volume reflects therefore a larger variability of lamellar body size (Fig. 1e and f). The total volume of lamellar bodies per lung was increased threefold in A−D− mice. The variation analysis of all parameters revealed a minor contribution of the variation introduced by the stereological sampling design to the total observed variation (data not shown). Therefore, the precision of the stereological estimates was sufficient to ensure that the total observed variation was dominated by the biological variation between animals.
Table 1. Stereological data
Data are means (CV) of n = 5 wild type (WT) as well as surfactant protein A and D double deficient (A-D-) mice.
Abbreviations of parameters: N, number, νN, number- weighted mean volume; νv, volume-weighted mean volume; V, total volume; S, surface area; alv, alveoli; type II, alveolar type II cells; AM, alveolar macrophages; nucl, nuclei; mito, mitochondria; lb, lamellar bodies; basolat, basolateral. The reference volumes are either the total lung volume (lung) or the number-weighted mean volume of alveolar type II cells (type II).
The lung collectins, SP-A and SP-D, potentially contribute to the biophysical (alveolar stability) as well as the immunomodulatory (alveolar sterility) surfactant functions (Crouch and Wright, 2001; Hawgood and Poulain, 2001; McCormack and Whitsett, 2002). The present design-based stereological analysis characterizes parenchymal, alveolar type II cell, and alveolar macrophage alterations and thus provides quantitative data reflecting that chronic deficiency of SP-A and SP-D in mice leads to parenchymal remodeling, type II cell hyperplasia and hypertrophy, accumulation of enlarged alveolar macrophages, and disturbed type II cell surfactant metabolism. Since this study was performed using mice of B6/129Sv:CD-1 mixed genetic background (Hawgood et al., 2002), some variability due to differences in genetic background is possible. Littermates were used to minimize this variability.
The recently established unbiased estimation of alveolar number and size (Hyde et al., 2004; Ochs et al., 2004a) provides efficient parameters to compare lung morphology objectively, e.g., when different genotypes with emphysematous alterations are studied (Ochs et al., 2004b). The decrease in alveolar number together with an increase in mean alveolar size in A−D− mice underlines emphysema-like pathology (Hawgood et al., 2002). The extent of the emphysematous alterations is comparable to SP-D single deficient mice (Ochs et al., 2004b). However, direct comparisons await the analysis of fully backcrossed littermates. The pathological mechanisms that lead to the parenchymal remodeling in lung collectin-deficient mice remain to be further investigated.
The quantitative characterization of hyperplastic and hypertrophic alveolar type II cells by stereological means is preferable to qualitative microscopy, because objective comparisons are possible. The proliferation of type II cells in A−D− mice might be a reaction to lung inflammation. In different models of acute and chronic lung inflammation, type II cell proliferation and hypertrophy were demonstrated (Miller and Hook, 1990; Bryson et al., 1991; Kasper and Haroske, 1996; Fehrenbach, 2001; Homer et al., 2002). In human lung emphysema, the rate of proliferation of alveolar epithelial cells is enhanced (Yokohori et al., 2004). However, recent studies on the effects of an additional loss of GM-CSF in SP-D-deficient mice indicate that type II cell and intracellular surfactant alterations may not be causally related to the chronic inflammation and emphysema in SP-D-deficient mice (Ochs et al., 2004b).
The occasional enlargement of single lamellar bodies in SP-D-deficient mice does not lead to an increase in mean lamellar body size. However, it is reflected in an increased volume-weighted mean volume, thus indicating an increased variability in size. By using unbiased sampling techniques, stereology assesses the whole organ, not only the most severely affected regions focused by the researcher's eye. The present study demonstrates an elevated total volume of lamellar bodies per lung due to an increase in the number of type II cells and in the number of lamellar bodies per cell. This accumulation of intracellular surfactant, together with the progressive accumulation of intra-alveolar surfactant that was demonstrated previously (Hawgood et al., 2002), reflects metabolic disturbances either directly due to A−D− deficiency or connected with lung inflammation (Lesur et al., 1995; Viviano et al., 1995; Fehrenbach et al., 1998; Ikegami et al., 2000b). At present, it is not clear if the lung collectins directly influence type II cell proliferation and metabolism or rather indirectly by modulating lung inflammation in vivo. Recently, SP-A has been shown to influence surfactant metabolism under challenging conditions such as hyperventilation (Jain et al., 2003). A stereological approach to quantify lung structural alterations might be helpful to study A−D− mice challenged by various stimuli.
In conclusion, chronic deficiency of the lung collectins, SP-A and SP-D, in mice leads to emphysema-like pathology with fewer and larger alveoli, type II cell hyperplasia and hypertrophy, and an increased intracellular surfactant pool. The pathomechanisms responsible for these morphological alterations await further examination. The present study also underscores the value of quantitative morphology for the phenotype characterization of mutant mice. The design-based stereological approach presented here provides a framework for the quantitative lung structure analysis in gene-manipulated mice as well as potentially for the study of humans with surfactant deficiencies or other lung diseases.
Gratitude to S. Freese, A. Gerken, H. Hühn (Göttingen), C. Brown, J. Edmondson, and D. Ansaldi (San Francisco) for technical assistance and to C. Maelicke for checking the article. Supported by the Medical Faculty of the University of Göttingen (1400490; to M.O.); the Alexander von Humboldt Foundation Feodor Lynen Fellowship (to M.O.), and the National Institutes of Health (HL-24075 and HL-58047; to S.H.)